MiNDEbEFF] SOUTHERN AND EASTERN ROOMS. 313 



there is no doubt that two stories above the ground were the maximum 

 height of the western rooms, exchiding the parapet. The eastern wall 

 presents a marked double convexity while the western wall is compara- 

 tively straight in a horizontal line, but markedly concave vertically 

 above the first roof level. Below this level it is straight. The floor 

 beams were from 3 to ti inches in diameter. The marks in the eastern 

 wail show that the beams projected into it to a nearly uniform depth 

 of 1 foot 4 inches. In the western wall, however, the depth varies from 

 1 to 3 feet. The beams which entered the eastern wall were veiy 

 irregularly placed, the line rising in the center some 3 or 4 inches. 

 The beams of the second roof level show the same irregularity and in 

 the same place; possibly this was done to correct a level, for the same 

 feature is repeated in the eastern room. 



The walls of the southern room are perhaps better finished and less 

 well constructed than any others in the building. The beam holes in 

 the southern wall are regular, those in the northern wall less so. The 

 beams used averaged a little smaller than those in the other rooms, and 

 there is no trace whatever in the overhanging wall of the use of rushes 

 or canes in the construction of the roof above. The walls depart con- 

 siderably from vertical plane surfaces; the southern wall inclines fully 

 12 inches inward, while in the northeastern corner the side of a door- 

 way pi'ojects fully 3 inches into the room. The broken condition of the 

 southern wall indicates carelessness in construction. The weakest 

 point in pise construction is of course the framing around openings. 

 In the southern wall the openings, being doubtless the first to give 

 way, are now almost completely obliterated. In the center of the wall 

 there were two openings, one above the other, but not a trace of lintels 

 now remains, and the eastern half of the wall now stands clear from 

 other walls. Probably there was also an opening near the southwestern 

 corner of the room, but the lintels giving way the waU above fell down 

 and, as shown on the ground plan (plate lii), filled up the opening. 

 This could happen only with exceptionally light lintels and exception- 

 ally bad construction of walls; one of the large blocks, before described 

 as composing the wall, must have rested directly above the opening, 

 which was practically the same size as the block. 



The walls of the eastern room were well finished, and, except the 

 western wall, in fairly good preservation. The floor beams were not 

 placed in a straight line, but rise slightly near the middle, as noted 

 above. The finish of some of the openings suggests that the floor was 

 but 3 or 4 inches above the beams, and that the roughened surface, 

 already mentioned, was not ])art of it. The northern wall of this room 

 seems to have run through to the outside, on the east, as though at 

 one time it formed the exterior wall of the structure ; and the eastern 

 wall of the building north of this room is separated from the rest of 

 the wall by a wide crack, as though it had been built against a smooth 

 surface. The western wall of this room shows clearly that in the con- 



