of
the building. At Luxor (fig. 86), the progress went on methodically enough
under Amenhotep III. and Seti I., but when Rameses II. desired to add to
the work of his predecessors, a bend in the river compelled him to turn
eastwards. His pylon is not parallel to that of Amenhotep III., and his
colonnades make a distinct angle with the general axis of the earlier work.
At Philae (fig. 87) the deviation is still greater. Not only is the larger
pylon out of alignment with the smaller, but the two colonnades are not
parallel with each other. Neither are they attached to the pylon with a due
regard to symmetry. This arises neither from negligence nor wilfulness, as
is popularly supposed. The first plan was as regular as the most
symmetrically-minded designer could wish; but it became necessary to adapt
it to the requirements of the site, and the architects were thenceforth
chiefly concerned to make the best of the irregularities to which they were
condemned by the configuration of the ground. Such difficulties were, in
fact, a frequent source of inspiration; and Philae shows with what skill
the Egyptians extracted every element of beauty and picturesqueness from
enforced disorder.

[Illustration: Fig. 88.--Plan of Speos, Kalaat Addah, Nubia.]

[Illustration: Fig, 89.--Plan of Speos, Gebel Silsileh.]

[Illustration: Fig. 90.--Plan of the Great Speos, Abu Simbel.]

[Illustration: Fig. 91.--Speos of Hathor, Abu Simbel.]

[Illustration: Fig. 92.--Plan of the upper portion of the temple of Deir el
Bahari, showing the state of the excavations, the Speos of Hathor (A); the
rock-cut sanctuary (B); the rock-cut funerary chapel of Thothmes I. (C);
the Speos of Anubis (D); and the excavated niches of the northern
colonnade. Reproduced from Plate III. of the _Archaeological Report of the
Egypt Exploration Fund_ for 1893-4.]

[Illustration: Fig. 93.--Plan of temple of Seti I., at Abydos.]

The idea of the rock-cut temple must have occurred to the Egyptians at an
early period. They carved the houses of the dead in the mountain side; why,
therefore, should they not in like manner carve the houses of the gods? Yet
the earliest known Speos-sanctuaries date from only the beginning of the
Eighteenth Dynasty. They are generally found in those parts of the valley
where the cultivable land is narrowest, as near Beni Hasan, at Gebel
Silsileh, and in Nubia. All varieties of the constructed temple are found
in the rock-cut temple, though more