n lighted in this manner, yet
the inner chambers must have remained in almost entire darkness. And
it is not improbable that such was the case, to judge from modern
Eastern houses, in which the rooms are purposely kept dark to mitigate
the great heat. The sculptures and decorations in them could then only
be properly seen by torchlight. The great courts were probably open to
the sky, like the courts of the modern houses of Mosul, whose walls
are also adorned with sculptured alabaster. The roofs of the large
halls must have been supported by pillars of wood or brick work. It
may be conjectured that there were two or three stories of chambers
opening into them, either by columns or by windows. Such appears to
have been the case in Solomon's temple; for Josephus tells us that the
great inner sanctuary was surrounded by small rooms, "over these rooms
were other rooms, and others above them, equal both in their measure
and numbers, and these reached to a height equal to the _lower part_
of the house, for the upper had no buildings about it." We have also a
similar arrangement of chambers in the modern houses of Persia, in
which a lofty central hall, called the Iwan, of the entire height of
the building, has small rooms in two or three separate stories opening
by windows into it, whilst the inner chambers have no windows at all,
and only receive light through the door. Sometimes these side chambers
open into a center court, as we have suggested may have been the case
in the Nineveh palaces, and then a projecting roof of woodwork
protects the carved and painted walls from injury by the weather.
Curtains and awnings were no doubt suspended above the windows and
entrances in the Assyrian palaces to ward off the rays of the sun.

Although the remains of pillars have hitherto been discovered in the
Assyrian ruins, we now think it highly probable, as suggested by Mr.
Fergusson, that they were used to support the roof. The modern Yezidi
house, in the Sinjar, is a good illustration not only of this mode of
supporting the ceiling, but of the manner in which light may have been
admitted into the side chambers. It is curious, however, that no stone
pedestals, upon which wooden columns may have rested, have been found
in the ruins; nor have marks of them been found on the pavement. We
can scarcely account for the entire absence of all such traces.
However, unless some support of this kind were resorted to, it is
impossible that the large