touched, while a
considerable space remained between their upper extremities instead of a
key-stone, or a key-brick fitting the aperture, ordinary bricks were
placed in it longitudinally, and so the space was filled in.

[Illustration: PLATE 59]

Another mode of constructing a pointed arch seems to be intended in a
bas-relief, whereof a representation has been already given. The masonry
of the arcade in No. V. [PLATE XLIX., Fig. 4] runs (it will be seen) in
horizontal lines up to the very edge of the arch, thus suggesting a
construction common in many of the early Greek arches, where the stones
are so cut away that an arched opening is formed, though the real
constructive principle of the arch has no place in such specimens.

With regard to the uses whereto the Assyrians applied the arch, it would
certainly seem, from the evidence which we possess, that they neither
employed it as a great decorative feature, nor yet as a main principle
of construction. So far as appears, their chief use of it was for
doorways and gateways. Not only are the town gates of Khorsabad found to
have been arched over, but in the representations of edifices, whether
native or foreign, upon the bas-reliefs, the arch for doors is commoner
than the square top. It is most probable that the great palace gateways
were thus covered in, while it is certain that some of the interior
doorways in palaces had rounded tops. Besides this use of the arch for
doors and gates, the Assyrians are known to have employed it for drains,
aqueducts, and narrow chambers or galleries. [PLATE LVIII. Fig. 2.];
[PLATE LIX., Fig. 1.]

It has been suggested that the Assyrians applied the two kinds of arches
to different purposes, "thereby showing more science and discrimination
than we do in our architectural works;" that "they used the pointed arch
for underground work, where they feared great superincumbent pressure
on the apex, and the round arch above ground, where that was not to be
dreaded." [PLATE LIX., Fig. 2.] But this ingenious theory is scarcely
borne out by the facts. The round arch is employed underground in two
instances at Nimrud, besides occurring in the basement story of the
great tower, where the superincumbent weight must have been enormous.
And the pointed arch is used above ground for the aqueduct and hanging
garden in the bas-relief (see [PLATE XLIX., Fig. 4]), where the
pressure, though considerable, would not have been very extraordinary.
It woul