we should lay gold leaf. It bears an inscription in cuneiform
characters.

[Illustration: FIG. 72.--Sheath of a cedar-wood mast, bronze.]

We are inclined to take these plates for models in restoring the columns of
the Sippara tabernacle. There is nothing in the richness of this double
covering of bronze and gold to cause surprise, as the inscription which
covers part of the face and the whole of the back of the tablet is nothing
but a long enumeration of the gifts made to the shrine of Samas by the
reigning king and his predecessors.

This column has both capital and base. The former cannot have been of
stone; a heavy block of basalt or even of limestone would be quite out of
place in such a situation. As for the base it is hardly more than a
repetition of the capital, and must have been of the same material; and
that material was metal, the only substance that, when bent by the hand or
beaten by the hammer, takes almost of its own motion those graceful curves
that we call _volutes_.

We believe then in a bronze capital gilded. Under the volutes three rings,
or _astragali_, may be seen. By their means the capital was allied to the
shaft. The former consisted of two volutes between which appeared a
vertical point resembling one of the angles of a triangle. The base is the
same except that it has no point, and that the rings are in contact with
the ground instead of with the shaft. These volutes may also be perceived
on the table in front of the tabernacle, where they support the large disk
by which the sun-god is symbolized.

[Illustration: FIG. 73.--Interior of a house supported by wooden pillars;
from the gates of Balawat. British Museum.]

Before quitting this tablet we may point to another difference between the
column of Sippara and the shafts of the same material and proportions that
we have encountered in the Assyrian bas-reliefs (Figs. 67, 68, and 69). In
the latter the column rises above the canopy, which is attached to its
shaft by brackets or nails. At Sippara the canopy rests upon the capital
itself. The same arrangement may be found in Assyrian representations of
these light structures; it will suffice to give one example taken from the
gates of Balawat (Fig. 73). Here, too, the proportions of the columns prove
them to have been of wood. They do not rise above the entablature. The
architrave rests upon them, and, as in Greece and Egypt, its immediate
weight is borne by abaci.

At present our aim is