rised within the area _b_
_c_ _d_ _d_, has fifteen rooms, or five longitudinal rows of three,
whereas the western has six rows of three. The rooms east must therefore
be larger than those west, and we see that they measure from east to
west respectively, 2.25 m., 2.28 m., and 2.28 m.--7 ft., 7 ft. 6 in.,
and 7 ft. 6 in.: from north to south, 3.60 m., 5.07 m., 4.43 m., 4.13
m., and 3.43 m.--12 ft., 17 ft., 15 ft., 14 ft., and 11 ft. It is a
rectangle, or rather trapezium, 22.31 m. x 7.81 m.--70 ft. x 25
ft.,--consequently the width of the building _B_ is somewhat less on the
line _d d d_ than on the line _a b c_. The cause of this singular
contraction I have found, and shall afterwards indicate.

Then follows a transverse section (_d d d e e_), containing two rows of
six rooms each, or twelve in all, of very unequal sizes, as the
ground-plans show. This entire section appears to be trapezoidal. The
line _d d d_ is 15.89 m.--52 ft.--long; the line _e e_ 16.33 m.--53 ft.;
_d e_ measures 7.42 m.--24 ft.--along the west, and 8.04 m.--27
ft.--along the east. Rooms marked _II_ and _III_ are particularly
irregular, having, as the diagram shows, not less than six corners.

From _e e_ to _f f_, another transverse section, this time of four rows
of six each, or twenty-four cells in all, those of each row being of
equal length, to wit 3.65 m.--12 ft.; and in width from east to west,
respectively: 2.25 m., 2.78 m., 3.18 m., 2.63 m., and 4.40 m.--7 ft., 9
ft., 10 ft., 9 ft., and 14 ft. (the last measure being the aggregate of
the two eastern compartments, the longitudinal partition being nearly
obliterated). To the south of _f f_ a further slight change occurs,
inasmuch as the three eastern rooms, instead of being respectively 2.68
m., 2.20 m., and 2.20 m.--9 ft., 7 ft., and 7 ft.,--now become 2.25 m.,
2.33 m., and 2.32 m.--7 ft., 8 ft., and 8 ft. From _f f_ to _g g_, the
southern limits of the structure, the whole structure is badly ruined;
and while the rooms can be counted, measurements are possible only in a
few places. Still I am satisfied that no great error lies in the
assumption that they were, taken longitudinally, all equal to the six
rooms contained in the transverse row south of the line _f f_, that is,
3.65 m.--12 ft.--from north to south; and in width, counting the cells
from west to east, respectively, 2.25 m., 2.78 m., 3.18 m., 2.25 m.,
2.33 m., and 2.32 m.--7 ft., 9 ft., 10 ft., 7 ft., 8 ft., and 8 ft. The
section,