From the appearance of the ground-plan, as I have been compelled to give it, it would result that the "first floor" contained two hundred and eleven cells, or rooms. Such is, however, not the case. The builders of this extensive fabric had not the means of preparing the hard rock foundation by removing it wherever it protruded over an average level. While giving a uniform height to their structure, they accommodated its ground-plan to the sinuosities of the rock. Out of this accommodation the irregularities noticed in the construction have mainly arisen. Pl. II., Figs. 1, 2, 3, will illustrate this statement. Pl. II., Fig. 1.--Cross-section of _B_ along the line a b c, north end; _a b_, actually visible top-line; _c d e f g h_, rock; _i k_, top of probable highest story, now destroyed. I have every reason to assume that this cross-section holds good for the entire division (_a b c d d_). From _d d_ on to _f f_ the distance between the rim of the _mesilla_ to the east and the house is greatest; the top-rock bends also to the west about _e e_, and there the irregularities noticed on the diagram about the chambers (_II_ and _III_) come in. They evidently result from an effort to conform the general plan to both the lateral and vertical deviations of its base. About the line _f f_, while the same number of chambers (six) remains in every transverse row, there is but one story below the general surface to the east. I may safely assume that south of the line _f f_ all the rooms of the first floor were on the same level. Pl. II., Figs. 2 and 3 will illustrate this point. As far as I could detect, the line _e e_ can be admitted as the one where one of the two lower stories disappears, and but one remains on the east side lower than the rest. [Illustration: PLATE II PLAN OF SECTIONS OF BUILDING B.]