and has disappeared without a trace, the restoration of this part of the
building is difficult. But the normal forms may yet be determined with
greater correctness than is presented in Coste’s restoration (_Fig._
82), which is a tasteful combination of the scotia and roundlet cornice
common to both Egyptian and Assyrian architecture, with dentils and the
leaved ornaments found above all the doors and windows of Persian
remains, and with the decorations upon the borders of staircase
buttresses. A number of rock-cut tombs appertaining to the early Persian
kings, the Achemenidæ, and dating from the time of Darius, represent the
façades of royal palaces, and give important information concerning the
exterior appearance of such structures. The oldest and best-preserved of
these is designated by cuneiform inscriptions as the tomb of Darius.
(_Fig._ 83.) It is especially interesting as illustrating the formation
of the entablature. An epistyle, triply stepped, like that of the Ionic
style, so that each face slightly projects beyond the one beneath it, is
placed above the transverse beam, which lies upon the backs of the
double-headed animals forming the capitals of the columns. The
multiplication of the faces of the epistyle is explained by the weakness
of the timber produced by Mesopotamia and Persia, which, in opposition
to the single and massive Doric lintel-block, required the employment of
several beams to obtain the desired capability of support. Upon it
followed the ornaments known as dentils, representatives of the small
and closely lying joists of the horizontal, slightly projecting roof.
They are quite similar to the dentils upon the tombs of Beni-hassan, and
to those of the still more naïve imitations of wooden houses found in
Lycia, which will be considered in the following section.
In Persia, the proportions of the dentils and of the distances between
them are still characteristic of the original timbered construction--a
truthfulness of imitation which was lost as early as the development of
the Ionic style. The nature of the band following above is not clear; it
might be natural to suppose in it a representative of such a hollow
cornice with leaves as Coste has introduced upon his entablature, were
it not that a frieze-relief with ornamental lions is visible upon this
member in another tomb, and that a remarkable block of the Palace of
Darius at Persepolis bears further testimony against it. One of the
corner piers of the front portico of that building has been preserved to
such a height that the side bearing of the lintel can be observed. This
renders the projection and outline of the entablature certain. It was
six times stepped, and may best be reconstructed, as in _Fig._ 84, by a
series of narrow bands, which represent in some measure the layers of
the horizontal ceiling and roof. From a comparison with the rock-cut
tomb, it is plain that a further cornice, like that over the door and
window-frames, was here not possible. If a parapet had been desired for
the accessible platform of the roof, it must have taken the form of a
light balustrade, not that of a heavy scotia cornice.
[Illustration: Fig. 84.--Entablature of the Palace of Darius.
Reconstructed from the Bearing.]
[Illustration: Fig. 85.--Plan of the Palace of Darius at Persepolis.]
The oldest and, because best-preserved, the most intelligible of the
royal dwellings upon the terrace of Persepolis is that shown by
inscriptions to have been built by Darius. (_Fig._ 85; and K upon the
topographical plan of Persepolis, _Fig._ 78.) It exhibits a regular and
well-considered plan, the oblong form and general disposition of which
are somewhat similar to the simpler Greek houses. A flight of steps led
from each side to the narrow southeastern front--a double tetrastyle
loggia. This was flanked by two moderately large rooms, which, as they
could be entered only from the portico and had no connection with the
interior, were probably intended for guards or servants. A door, between
four windows, opened into the square hall, the ceiling of which was
supported by sixteen columns, standing in line with those of the loggia.
This space corresponded to the atrium of Greek and Roman houses. Three
of its sides, that of the front being excepted, had access to inner
rooms--those upon the right and left being small, while, opposite the
entrance, they were more spacious, and separated from the hall by a
corridor. The walls were enriched by niches as well as by door and
window openings. Through one of the chambers upon the left was a
lateral entrance, reached by a double flight of steps upon the
southwest. Notwithstanding the preservation of the special foundation
terrace, of the steps, of the door, window, and niche frames, as well as
of some corner piers, the ruin did not at first glance make evident the
disposition here described. All the columns of the palace have
disappeared. It is uncertain whether this is because the supports of the
less pretentious structure were of wood, or whether stone shafts, of the
moderate dimensions which must be assigned to them, were carried away
during the two thousand years in which the ruins of the palace terrace
have served as a quarry for neighboring towns. The square plinths upon
which the columns stood have, however, remained in their original
position, so that the number and site of the supports may be easily and
surely determined. The greater portion of the walls has also
disappeared. Some corner piers and the marble frames of doors, windows,
and niches, cut from immense monolithic blocks, alone stand erect; but
their perfect state of preservation and well-marked position permit the
nature of the wall between them to be determined without difficulty. It
seems that this was of small quarried stones, or even of brick, thus
being easily removed, or, in the latter case, reduced to dust by
atmospheric influences; while the massive door and window casings were
secure from removal by man and from the injuries of time. Their stepped
jambs are decorated upon the inner side with reliefs; the heavy lintels
have a scotia cornice, carved with a triple row of leaves and bordered
below by an astragal. Of the openings for providing light to the great
hall no traces remain. If, as is usually supposed, the windows now
recognizable were all that ever existed, the chambers of the palace
would have been most gloomy, with the exception of the hall of columns,
which had four openings upon the loggia, besides the door. The light of
the hall itself must have been dim, for it could not enter directly, the
windows and doors being beneath the shade of the deep portico, with its
double range of columns; and when still more impeded by the
close-standing shafts of the hypostyle, it would have been wholly
insufficient for the chambers. It is further to be remarked that several
of the inner rooms have no direct communication with the hall, while if
they had depended on it for light they would certainly have been
provided with window-openings in place of the blind niches. It is
evident from the existence of a second story, presently to be discussed,
that horizontal apertures in the roof and ceiling could not have
existed; this would be even more inadmissible here than in the palace
buildings of Nineveh. It is necessary, however, to assume other openings
for illumination and ventilation than those now to be observed in the
ruins, and windows were most probably arranged in the manner in which
the Orientals still secure their dwellings from the view of the outer
world while admitting light and air--the manner customary with the
Assyrians, as well as with the more ancient Greeks. The apertures were
probably upon the exterior walls, just under the ceiling, high above the
ground. All traces of architectural members in such a position must
necessarily have disappeared when the mass of masonry which supported
them was overthrown. It is possible that their form was entirely plain,
like that given in the restoration of the Palace of Darius at the head
of this section (_Fig._ 77), and offered no carved details to aid in
their recognition.
[Illustration: Fig. 86.--Persian Door-casing.]
A comparison of the rock-cut façade upon the tomb of Darius with the
palace of that king will aid in the consideration of the upper story. As
the tomb represents the palace with but slight variations, even agreeing
tolerably well with its proportions, it may be supposed that the monarch
copied his dwelling upon the front of his grave, that he might, as it
were, inhabit it even after his death. This is not an isolated instance
of such a proceeding in the history of architecture. The second story,
distinctly recognizable upon the tomb, cannot be regarded as an
insignificant decoration, especially as the Palace of Darius at
Persepolis seems, from its plan, to have been thus arranged. The
limited area covered, exceeded by many a modern private house, renders
an enlargement by a second story natural; and this is also made probable
by the hypostyle, which occupies a place where an open court, with full
upper light, would otherwise have been more suitable. Space for the
staircases was provided by the two narrow corridors next the rear
chambers. The second story was not, however, extended over the entire
ground-plan, but seems to have left the flat roof of the side chambers
as an elevated veranda, perhaps sheltered from the sun by canopies, as
the _talar_, a similar though smaller upper structure, stands as a
pavilion upon the modern houses of Persia. The walls of the second story
could scarcely have been placed elsewhere than upon the otherwise
unreasonably thick partition-enclosure of the hypostyle hall. They could
not have stood over an intercolumniation, as upon the façade of the
rock-cut tomb--for this would have been difficult, if not impossible of
construction--but in other respects the upper part of the palace may
have been like that representation. Its corner supports, which are a
strange combination of scotias and roundlets, ending below in lion’s
paws and above in a one-sided lion capital, have, at least, every
appearance of being copied from an architectural model, and are similar
in their lower half to the legs of the throne given in _Fig._ 87. The
standing figures, which, in double row, support the ceiling, may have
been carved in relief or simply painted. That this was a common ornament
is evident from its repetition upon the reliefs of gateways, where such
typical figures are admirably characterized as representatives of the
various nations subjugated by the Persian power, they literally
supporting the throne. The entrance and the second-story windows may be
supposed to have been upon the side opposite the front, where the
veranda was broadest and the staircases led from the lower floor, as
otherwise the imitation of the façade upon the rock-cut tomb would have
shown windows and doors as well as a staircase, which probably led in
double flight to the uppermost roof. That this house-top was flat and
accessible is evident from the reliefs considered in this connection
(_Figs._ 83 and 87), one of which represents the royal throne shaded by
a canopy, the other one of those fire-altars which, according to Persian
custom, was placed upon the highest level of the house. This altar upon
the summit of a royal palace is mentioned in the Bible, when Hezekiah,
overthrowing the Sabæan worship of the sun, destroyed “the altar which
is upon the top of the upper rooms of Ahaz.” In the restoration of the
Palace of Darius (_Fig._ 77), the introduction of the altar with the
royal canopy may be considered as more than a mere decoration of the
design. This simplest and best-preserved ruin upon the terrace of
Persepolis permits a comparatively trustworthy understanding of the
elements of Persian palace architecture.
[Illustration: Fig. 87.--Relief from the Portal of the Hall of Hundred
Columns.]
The ruin O of the topographical plan (_Fig._ 78) shows the remains of a
similar structure of about the same dimensions, later, and therefore of
less interest, than the Palace of Darius. The Palace of Xerxes (L, M, N)
was nearly double this size, being provided with a spacious terrace
before its gates, and with a colonnade upon one side, the nature of
which cannot readily be explained. On the other hand, it had no large
chambers behind the hypostyle, as the rooms upon the right and left
seem, by their more spacious proportions, to have rendered these
unnecessary. The portico was hexastyle, and the square hall behind it
consequently provided with thirty-six columns. Two of the side chambers
were so large that their ceilings required the support of four columns.
Of still greater dimensions, more than eight times the area of the
Palace of Darius, was the Palace Hall of Xerxes (D, E, F, G) which was
preceded by a magnificent double flight of steps. The ceiling of the
imposing hypostyle was upheld by thirty-six columns of gigantic size.
There are no traces of chambers having been connected with it; three of
its sides were provided with hexastyle porticos, which masked and
artistically enlivened the dead enclosing-walls. The masonry has
disappeared, with the exception of unimportant remains of the portal
(G), which Coste has restored as the foundations of pedestals. Although
a similar ruin at Susa, examined by Loftus, was also without walls, it
is impossible to agree with Coste that these were originally altogether
lacking, and that the columns of the central space were unenclosed--that
the three portals, provided with separate roofs, were grouped around
this without any connection. While we agree with Fergusson in as far as
regards the completion of the wall line and the unity of the whole under
a common roof, we must yet discredit his further assumption that this
building was provided, like the Palace of Darius, with an upper story;
all the requisite conditions for this were lacking. The ruin is
remarkable from the remains of the colossal columns being in the
comparatively best state of preservation. They represent the three
orders described above: those of the western portico having the
double-headed bull; those of the eastern the double-headed lion, and the
others the form of shaft coronation combined of three or four members.
The destination of this building was not that of a dwelling, but,
without doubt, that of a festive hall for the audiences and ceremonies
of the vainest and most magnificent of despotic monarchs. To this end it
was fittingly placed next to the entrance-gate of the palace terrace. It
is one of the most enormous buildings of the world; the area covered by
its plan, about 10,500 sq. m., nearly equals that of the Cathedral of
Milan, and surpasses that of the Cologne cathedral by about 2350 sq. m.
[Illustration: Fig. 88.--Propylæa of Xerxes at Persepolis.]
The imposing portal next to it, B, proved by inscriptions to have also
been erected by Xerxes, remains upright in the grand masses shown by
_Fig._ 88. An adequate explanation of its nature is not possible. It is
only clear that its principal disposition, like that of the similar
portal, H, of the terrace, was determined by the intersection of
passages, the crossing being marked by four columns, while the parallel
walls were of sculptured marble blocks. In a former work upon the
history of ancient architecture,[F] the author has expressed the
supposition that side walls were built in the directions marked by
dotted lines upon the topographical plan (_Fig._ 78), connecting the
portal with the ascending staircase. The gate would thus receive the
character of a fortification, a termination of the palace terrace,
instead of being the useless structure, easily to be circumvented, which
it is commonly considered. It is probable that these side walls existed
also at the chief portals of the Assyrian palaces, as otherwise the
entrances, especially that of the harem, would have been too much
exposed. These masses of masonry have disappeared from the ruins of
Nineveh, because of the crumbling of the terrace borders, and in
Persepolis, where all walls have been overthrown and carried away, their
extent is not marked by the more durable door and window frames, which
alone remain of the palace enclosures.
The assumption of similar communicating walls in connection with the
other portal structures of the palace terrace (H and Q) not only renders
to these their full importance, but throws light upon a building of
enormous extent (C), the destination of which has hitherto been
problematical. This edifice has been called, in lack of a better name,
the Hall of a Hundred Columns. It is an extended enclosure of square
plan, within which stood columns, traceable by the remains of six of
their number. Upon the front was a portico, not decastyle, like the
interior, but octastyle; two bases remaining _in situ_ determine its
arrangement and dimensions. The columns may be calculated, from their
lower diameter, to have been about 7 m. high. The enclosure of the hall,
determined in extent by the remains of all the portals and niches,
measured 68 m. upon each side. According to general acceptance, the
building was restricted to the area now covered by its ruins, and served
as a second great hall for ceremonies. Fergusson terms it a coronation
hall. But, apart from the fact that the Hall of Xerxes must have been
far better fitted by its imposing proportions for such a purpose than
this low and broad space, where the forest of columns would have impeded
the view, it is hardly possible that two such extensive buildings would
have been provided upon the terrace for the same use. But some adequate
space is yet to be assigned to that important necessity of Oriental
custom, the harem, which tradition particularly asserts to have existed
among the Persian palaces. If the ruin is examined in its relation to
the other palace structures of Persepolis, it becomes plain that it can
be nothing else than the central hall of a similar, but more extended,
series of chambers, of which, as is also the case with the ruined
remains at O, hypostyle and portico have alone been preserved, while the
walls of all the outer rooms have disappeared. Only the doors and
windows of any wall upon the terrace now exist; and as the entrances
were naturally small and the openings for light high above the ground,
in the enclosure of the harem, it is not surprising that this masonry
has disappeared in almost its entire extent. Two principal portals,
perhaps the only ones of the outermost walls, have been preserved,
however, and mark the outline of the building. These are the gateways H
and Q of the topographical plan: the first of these even shows some
trace of the enclosing wall; it is the entrance from the palaces K, L,
M, N, and O; the second probably led to an open court, to which access
must have been allowed the fair prisoners. The space between the
hypostyle and the exterior wall, indicated upon the plan by dotted
lines, must have been occupied by the numerous small rooms which
provided dwellings for the three hundred girls of the harem. The low and
broad central hall served as a place of assemblage; the great number of
its columns and the excessive lowness of the ceiling exclude the idea of
its having been used for public ceremonies, but render it particularly
fitted for this purpose, the many shafts separating the groups of
intimate conversers. The dim twilight of the room was, at these evening
assemblies, enlivened by the many-colored lamps of the East. The harem
upon the terrace thus received a development analogous to that of the
royal dwellings, and its necessarily great extent was provided for in a
becoming place. By the assumption that the remains at P are those of the
harem, an integral part of the Oriental palace is recognized, and a
large tract of the terrace area is occupied, the use of which could not
otherwise be designated upon the topographical plan.
The disposition of the terrace under Darius appears to have differed
considerably from that under his successors. It is not known whether its
extent has since been increased; to establish this point, extensive
excavations would be required. It is probable that the northwestern side
of the plateau has been built out by adding earth to the natural rock;
the buildings upon the southern half appear the more primitive: it is
certain, however, that the position of the ascent was changed during the
great reconstruction completed by Xerxes, and possibly commenced during
the latter part of the reign of Darius. The orientation of the Palace of
Darius, which, of all the buildings at Persepolis, alone faces the
south, shows the great staircase to have been originally upon the
southern end of the terrace. Enormous dowelled blocks of stone assured
the stability and preservation of the newer parts of the substructure.
The broad and gently rising flights of steps remain in so good a
condition that it is even to-day possible to ascend them upon horseback.
[Illustration: Fig. 89.--Altar Pedestals at Pasargadæ.]
Among the remaining monuments of Persian architecture there are no
temples; it would be vain to seek such structures; the worship of the
land did not demand closed rooms, requiring only sacrifice and prayer
upon the summits of mountains or artificial elevations. Herodotos
relates that the Persians not only scorned temples, but did not erect
images of their deities, nor even altars. This last point is certainly
incorrect; the worship of fire particularly called for altars, and these
are represented upon the ornamented façades of the rock-cut tombs.
(_Fig._ 83.) It is probable that two pedestals, standing near each other
upon the palace terrace of Pasargadæ, are ancient Persian. They are
cubes, each about 3 m. high; one is terminated by steps, and has upon
one side a straight line of ascending stairs; the platform at the summit
was sufficiently large to receive an altar, or may perhaps itself have
been used as a receptacle for fire and sacrifices. They are similar to
the altar upon the upper story of the Palace of Darius, used for
religious devotion. The supposition may be ventured that these two
altars, in such vicinity, point to the dualism of the Persian worship of
Ormuzd and Ahriman.
Other large monuments of the land may have had something to do with
religious observances; but as they lack any characteristic form, this
cannot be proved. Such is the case with the cone of Darabgerd, known as
Kella Darab, apparently an imitation of a natural mound. It is
surrounded by a circular wall, perforated in eight equidistant places,
and rises, in two rings of masonry, to a height of 48 m. A similar
structure is the massive tower of Firuz-Abad, a rectangular obelisk 27
m. high, measuring 8.5 m. upon each side of its base. Near it is an
enormous platform, with broad buttresses upon the four sides, which are
directed to the cardinal points of the compass; the foundation of the
mass measures 61 by 78 m. The masonry is of carefully hewn stone, of a
workmanship not found in the country after the advent of the Christian
era; the swallow-tail dowelling of the blocks is similar to that upon
the pavement of the terrace at Persepolis.
[Illustration: Fig. 90.--Tomb of Cyrus.]