The exploration of the Assyrian palaces has brought three systems of flooring to light—beaten earth, brick pavements, and pavements of limestone slabs.[297] In the palace of Sargon nearly every chamber, except those of the harem, had a floor of beaten earth, like that in a modern fellah's house. Even the halls in which the painted and sculptured decoration was most sumptuous were no exceptions to this rule. There is nothing in this, however, to surprise those who have lived in the East; like the Turks, Arabs, and Persians of our own time, the Chaldæans and Assyrians were shod, except when fighting or hunting, with those babooshes or sandals that are so often figured in the bas-reliefs. These must have been taken off, as they are to-day, before entering a temple, a palace, or a harem. Moses was required to take off his shoes before approaching the burning bush, because the place on which he stood was holy ground. In the houses of their gods, in those of their kings and rich men, the floor would be covered with those rich carpets and mats that from one end of the East to the other conceal from sight the floors of white wood or beaten earth. In summer the mats are fresh and grateful to the bare feet, in the winter the carpets are soft and warm. The floors themselves are hardly ever seen, so that we need feel no surprise at their being left without ornament. So, too, was it in all probability in the palaces of Sargon and of other kings, and in the sacred buildings.

Elsewhere, however, we find a pavement constructed with the most scrupulous care, and consisting of three distinct parts,—two layers of large bricks with a thick bed of sand interposed between them. The lower course of bricks is set in a bed of bitumen which separates it from the earth and prevents any dampness passing either up or down. This system of paving was used in most of the harem chambers at Khorsabad as well as in the open courts and upon the terraces. Lastly, in certain rooms of the seraglio and harem, in a few of the courts, in the vestibules, before the gates of the city, and in paths across wide open spaces, a limestone pavement has been found. Wherever this pavement exists, the stones are of the same kind and placed in the same manner. The limestone is exactly similar to that in the retaining walls described on page 147. The stones are often more than three feet square, and from two feet six inches to two feet ten inches thick. Their shape is not that of a regular solid; it is more like a reversed cone, the base forming the pavement and the narrow end being buried in the ground. These stones are simply placed side by side without the use of mortar or cement of any kind, but their weight and peculiar shape gave a singular durability to the pavement for which they were used.

Most of the sills belong to this class. And in Assyria where doorways were several yards deep and two or three wide, these sills were in reality the pavements of passages or even chambers.[298]

The materials for these pavements were always different from those of the floors on each side of them. In the entrances to the brick-paved courts large stones were used; in the passages between rooms floored with beaten earth bricks were introduced. The stone thresholds were mostly alabaster like the sculptured slabs upon the chamber walls. As a rule they were of a single piece, the great extent of surface, sometimes as much as ten or eleven square yards, notwithstanding. In the entries flanked by the winged bulls the sills were carved with inscriptions, which were comparatively rare elsewhere. Sometimes we find a rich and elaborate ornamentation in place of the wedges; it is made up of geometrical forms and conventional foliage and flowers; the figures of men and animals are never introduced. Such an arrangement was in better taste than the mosaic thresholds of the Romans where men were shown in pictures destined to be trodden under foot. The Assyrian carver doubtless took his designs from the carpets in the adjoining chambers.

A good idea of these designs may be formed from the slab figured below. The centre is occupied by a number of interlacing circles, betraying no little skill on the part of the ornamentist. The "knop and flower" border of alternately closed and shut lotus flowers is separated from the centre by a band of rosettes. The whole is distinguished by thought and a severe taste. The indented corners, where the pivots of the doors were placed, and the slot for the lower bolt of the door near the centre, should be noticed. These details prove that in this instance the door was a double one. In other cases the absence of the slot and the presence of only one pivot hole show that single doors were also used.[300] The doors always opened inwards, being folded back either against the sides of the entry itself or against the walls of the chamber.

Many of these sills or thresholds show no sign of a pivot at either corner, whence we may conclude that many of the openings were left without doors, and could only have been closed by those suspended carpets or mats of which such ready use is made in hot countries.

In very magnificent buildings metal thresholds sometimes replaced those of stone or brick. In the British Museum there is a huge bronze sill that was found in a ruined temple at Borsippa, by Mr. Rassam. Its extreme length is sixty inches, its width twenty, and its thickness about three and a half inches. It bears an inscription of Nebuchadnezzar the arrangement of which proves that the sill when complete had double its present length, or about ten feet. Its upper surface is decorated with large rosettes within square borders. We need hardly say that it is a solid casting, and that its weight is, therefore, by no means trifling. The workmen who put in place and those who cast it must both have thoroughly understood what they were about. Even now, we are told, the latter operation would be attended by some difficulty.[301]

The founders who produced this casting could have no difficulty over the other parts of the door-case, and we have no reason to doubt the statement of Herodotus, who thus ends his account of how the walls of Babylon were built: "The walls had a hundred gates, all of bronze; their jambs and lintels were of the same material."[302]

These lintels and jambs must have been, like the Borsippa threshold, of massive bronze, or they would soon have been crushed by the weight they had to support. On the other hand, had doors themselves been entirely of that metal it would have been very difficult if not impossible to swing them upon their hinges, especially in the case of city gates like those just referred to. It is probable, then, that they were of timber, covered and concealed by plates of bronze. Herodotus indeed narrates what he saw, like a truthful and intelligent witness, but he was not an archæologist, and it did not occur to him when he entered the famous city which formed the goal of his travels, to feel the shining metal and find out how much of it was solid and how much a mere armour for a softer substance behind.

From fragments found at Khorsabad, M. Place had already divined that the Assyrians covered the planks of their doors with bronze plates, but all doubts on the point have been removed by a recent discovery, which has proved once for all that art profited in the end by what at first was nothing more than a protection against weather and other causes of deterioration. In 1878 Mr. Hormuzd Rassam, the fellow traveller of Sir Henry Layard, found in the course of his excavations in Assyria for the British Museum, some metallic bands covered with repoussé reliefs and bearing the name of Shalmaneser III. (895-825). The site of this discovery was Balawat, an artificial mound about fifteen miles to the east of Mossoul.[303] As soon as these bands had been examined in London by competent archæologists, they were recognized as having belonged to the leaves of a wooden door, which must have been nearly twenty-seven feet high and about three inches thick. This latter dimension has been deduced from the length of the nails used to keep the bands in place. At one end these bands were bent with the hammer round the pivot to which each half of the door was attached. These pivots, judging from the bronze feet into which they were "stepped," were about twelve inches in diameter.

It is easy to see from their shape how these feet were fixed and how they did their work (Fig. 97). The point of the cone was let into a hollow socket prepared for it in a block cut from the hardest stone that could be found. Such a material would resist friction better and take a higher polish than brick, so that it was at once more durable and less holding. Sockets of flint, basalt, trachyte, and other volcanic rocks have been found in great numbers both in Assyria and Chaldæa.[304] Instances of the use of brick in this situation are not wanting,[305] however, and now and then the greenish marks left by the prolonged contact of metal have been discovered in the hollows of these sockets.[306]

More than one method was in use for fixing the pivots of the doors and enabling them to turn easily. Sir Henry Layard brought from Nimroud four heavy bronze rings which must have been used to supplement these hollow sockets.[308] In one way or another bronze occupied a very important place in the door architecture of the Assyrians. In those cases where it neither supplied the door-case nor ornamented its leaves, it was at least used to fix the latter and to enable them to turn.

In Assyrian façades doors had much greater importance than in those architectural styles in which walls are broken up by numerous openings. Their great size, their rich and varied ornamentation, the important figures in high relief with which the walls about them were adorned, the solemn tints of bronze lighted up here and there by the glory of gold, the lively colours of the enamelled bricks that formed their archivolts, and finally the contrast between the bare and gleaming walls on either side and their depths of shadow—all these combined to give accent to the doorways and to afford that relief to the monotony of the walls of which they stood in so great a need. For Assyrian mouldings are even poorer than those of Egypt. The softness of crude brick, the brittle hardness of burnt brick, are neither of them well disposed towards those delicate curves by which a skilful architect contrives to break the sameness of a façade, and to give the play of light and shadow which make up the beauty of a Greek or Florentine cornice.

The only mouldings encountered in Assyria have been found on a few buildings or parts of buildings in which stone was employed. We may quote as an instance the retaining wall of the small, isolated structure excavated by Botta towards the western angle of the Khorsabad mound, and by him believed to be a temple.[309] The wall in question is built of a hardish grey limestone, the blocks being laid alternately as stretchers and headers. The wall is complete with plinth, die and cornice (Figs. 98 and 99). The latter is a true cornice, composed of a small torus or bead, a scotia, and a fillet. The elements are the same as those of the Egyptian cornice, except in the profile of the hollow member, which is here a scotia and in Egypt a cavetto, to speak the language of modern architects. The Egyptian moulding is at once bolder and more simple, while the vertical grooves cut upon its surface give it a rich and furnished aspect that its Assyrian rival is without.[310]

We have another example of Assyrian mouldings on the winged sphinx found by Layard at Nimroud (Fig. 85)—the sphinx, that is, that bore a column on its back. In section this moulding may be compared to a large scotia divided into two cavettos by a torus. Its effect is not happy. The Assyrians had too little experience in stone-cutting to enable them to choose the most satisfactory proportions and profiles for mouldings.

We may also point to the entablatures upon the small pavilions reproduced in our Figs. 41 and 42. They are greatly wanting in elegance; in one especially—that shown in Fig. 42—the superstructure is very heavy in proportion to the little temple itself and its columns.

The only moulding, if we may call it so, borrowed by Assyria from Chaldæa, and employed commonly in both countries, is a brick one. Loftus was the first to point it out. He discovered it in the ruined building, doubtless an ancient temple, in the neighbourhood of Warka, and called by the natives Wuswas. This is his description:—"Upon the lower portion of the building are groups of seven half-columns repeated seven times—the rudest perhaps which were ever reared, but built of moulded semicircular bricks, and securely bonded to the wall. The entire absence of cornice, capital, base or diminution of shafts, so characteristic of other columnar architecture, and the peculiar and original disposition of each group in rows like palm logs, suggest the type from which they sprang."[311]

With his usual penetration, Loftus divines and explains the origin of these forms. The idea must have been suggested, he thinks, by the palm trunks that were used set closely together in timber constructions, or at regular intervals in mud walls. In either case half of their thickness would be visible externally, and would naturally provoke imitation from architects in search of ornament for the bald faces of their clay structures.[312]

As to the effect thus obtained, the rough sketch given by Loftus hardly enables us to decide (see Fig. 100). From Assyria, however, come better materials for a judgment. We there often find these perpendicular ribs, generally in groups of seven, in buildings that have been carefully studied and illustrated upon a sufficient scale. We give an example from one of the harem gates at Khorsabad (Fig. 101), by which we may see at once that an ornamental motive of no little value was afforded by these huge vertical reeds with their play of alternate light and shadow, and the happy contrast they set up between themselves and the brilliant hues of the painted walls and enamelled bricks. The whole had a certain elegant richness that can hardly be appreciated without the restoration, in every line and hue, of the original composition.

Both at Warka and in the Khorsabad harem, these vertical ribs are accompanied by another ornament which may, perhaps, have been in even more frequent use. We mean those long perpendicular grooves, rectangular in section, with which Assyrian and Chaldæan walls were seamed. In the harem wall these grooves flank the group of vertical reeds right and left, dividing each of the angle piers into two quasi-pilasters. At Warka they appear in the higher part of the façade, above the groups of semi-columns. They serve to mark out a series of panels, of which only the lower parts have been preserved. The missing parts of the decoration may easily be supplied by a little study of the Assyrian remains. The four sides of the building at Khorsabad, called by M. Place the Observatory, are decorated uniformly in this fashion. The general effect may be gathered from our restoration of one angle. The architect was not content with decorating his wall with these grooves alone; he divided it into alternate compartments, the one salient, the next set back, and upon these compartments he ploughed the long lines of his decoration. These changes of surface helped greatly to produce the varied play of light and shadow upon which the architect depended for relief to the bare masses of his walls. The most ordinary workmen could be trusted to carry out a decoration that consisted merely in repeating, at certain measured intervals, as simple a form as can be imagined, and, in the language of art as in that of rhetoric, there is no figure more effective in its proper place than repetition.

The necessity for something to break the monotony of the brick architecture was generally and permanently felt, and in those Parthian and Sassanide periods in which, as we have said, the traditions of the old Chaldæan school were continued, we find the panel replaced by wall arcades in which the arches are divided from each other by tall pilasters. In general principle and intention the two methods of decoration are identical.

The Egyptian architect had recourse to the same motive, first, in the tombs of the Ancient Empire for the decoration of the chamber walls in the mastabas; secondly, for the relief of great brick surfaces. The resemblance to the Mesopotamian work is sometimes very great.[313]

We have explained this form by one of the transpositions so frequent in the history of architecture, namely, a conveyance of motives from carpentry to brickwork and masonry.[314] In the former the openings left in the skeleton are gradually filled in, and these additions, by the very nature of their materials, most frequently take the form of panels. The grooves that define the panels in brick or stone buildings represent the intervals left by the carpenter between his planks and beams. They could also be obtained very easily upon the smooth face of beams brought into close contact, either by means of the gouge or some other instrument capable of cutting into the wood. We may safely assert that in Chaldæa and Assyria, as in Egypt, it was with carpentry that the motive in question originated.

On the other hand, if there be a form that results directly from the system of construction on which it is used, that form is the crenellation with which, apparently, every building in Mesopotamia was crowned.[315]

The Assyrian brickwork in which so many vast undertakings were carried out consists of units all of one dimension, and bonded by the simple alternation of their joints. Supposing a lower course to consist of two entire bricks, the one above it would be one whole brick flanked on either side by a half brick. An Assyrian wall or building consists of the infinite repetition of this single figure. Each whole brick lies upon the joint between two others, and every perpendicular wall, including parapet or battlement, is raised upon this system.

Far from being modified by the crenellations, this bond regulates their form, dimensions, and distribution. The crenellations of the palace walls consist of two rectangular masses, of unequal size, placed one upon the other. The lower is two bricks'-length, or about thirty-two inches, wide, and the thickness of three bricks, or about fourteen inches, high. The upper mass equals the lower in height, while its width is the length of a single brick, or sixteen inches. The total height of the battlement, between twenty-eight and twenty-nine inches, is thus divided into two masses, one of which is twice the size of the other (see Fig. 104). The battlements are all the same, and between each pair is a void which is nothing but the space a battlement upside down would occupy. Fill this space with the necessary bricks, and a section of wall would be restored identical in bond with that below the battlements, with the one exception that the highest block of the battlement, being only one brick wide, is formed by laying three whole bricks one upon the other.[316]

The crenellations we have been describing are those upon the retaining walls of Sargon's palace at Khorsabad. Those of the Observatory are slightly different in that they are three stories high instead of two (Fig. 105). The lowest is three bricks wide, the second three, the topmost two. They are each three bricks high. Why were these battlements given a height beyond those of the royal palace? That question may be easily answered. The crenellations of the observatory were destined for a much more lofty situation than those of the palace. The base of the former monument rose about 144 feet above the summit of the artificial hill upon which it was placed; the total elevation was about 190 feet, a height at which ordinary battlements, especially when for the most part they had nothing but the face of the higher stories to be relieved against, would be practically invisible.

Whether composed of two or three stages this battlement was always inscribed within an isosceles triangle; in fact, when a third story was added, the height and the width at the base increased in the same proportions. M. Place lays great stress upon this triangle. He makes it cut the upper angles of each of the superimposed rectangles, as we have done in our Figs. 104 and 105, and he points out how such a process gives an outline similar to that of a palisade cut into points at its summit, a precaution that is often taken to render the escalade of such an obstacle more difficult, and M. Place is inclined to think that the idea of these crenellations was suggested by those of a wooden palisade, a succession of rectangles being substituted for a triangle in order to meet the special conditions of the new material. To us, however, it hardly appears necessary to go back to the details of wooden construction to account for these forms. We find no sign of M. Place's spiked palisades in the bas-reliefs. The inclosures of the Mesopotamian fields must have consisted of palm trunks and strong reeds; planks were hardly to be cut from the trees of the country. Moreover, the mason and bricklayer saw the forms of these battlements repeated by their hand every instant. Whenever they began a fresh course the first brick they placed upon the joint between two units of the course below was the first step towards a battlement. The decoration obtained by the use of these battlements was not a survival from a previous form, it was a natural consequence from the fundamental principle of Assyrian construction.

It has been thought that some of the buildings represented on the bas-reliefs have triangular denticulation in place of the battlements figured on the last page;[317] and there are, in fact, instances in the reliefs of walls denticulated like a palisade (see Fig. 38), but these must not, we think, be taken literally. In most cases the chisel has been at the trouble to show the real shapes of the battlements (Fig. 42), but in some instances, as in this, it has been content to suggest them by a series of zig-zags. Here and there we may point out a picture in stone which forms a transition between the two shapes, in Fig. 41 for example. Such an abbreviation explains itself. It is, in fact, nothing more than an imitation of the real appearance of the rectangular battlements when seen from a distance.[318]

The architect was not content with the mere play of light and shade afforded by these battlements. He gave them a slight salience over the façade and a polychromatic decoration. About three feet below the base of the crenellations the face of the wall was brought forward an inch or two, so that the battlements themselves, and some eight or ten courses of bricks below them, overhung the façade by that distance, forming a kind of rudimentary cornice (see Fig. 106). In very elaborate buildings enamelled bricks were inserted between the battlements and this cornice. These were decorated with white rosettes of different sizes upon a blue ground. The explorers of Khorsabad encountered numberless fragments of these bricks and some whole ones in the heaps of rubbish at the foot of the external walls. Their situation proved that they had come from the top of the walls, and on the whole we may accept the restoration of M. Thomas, which we borrow from the work of M. Place, as sufficiently justified (Fig. 106).[319]

This method of crowning a wall may seem poor when compared to the Greek cornice, or even to that of Egypt, but in view of the materials with which he had to work, it does honour to the architect. The long band of shadow near the summit of the façade, the bands of brilliantly coloured ornament above it, and the rich play of light and shade among the battlements, the whole relieved against the brilliant blue of an Eastern sky, must have had a fine effect. The uniformity from which it suffered was a defect common to Mesopotamian architecture as a whole, and one inseparable from the absence or comparative disuse of stone. But in the details we have been studying we find yet another illustration of the skill with which these people corrected, if we may so phrase it, the vices of matter, and by a frank use of their materials and insistence upon those horizontal and perpendicular lines which they were best fitted to give, evolved from it an architecture that proved them to have possessed a real genius for art.

The Assyrians seem to have been so pleased with these crenellations that they placed them upon such small things as steles and altars. In one of the Fig. 107.—Altar; from Rawlinson. Fig. 107.—Altar; from Rawlinson. Kouyundjik reliefs (Fig. 42) there is a small object—a pavilion or altar, its exact character is not very clearly shown—which is thus crowned. Another example is to be found in a bas-relief from Khorsabad (Fig. 107).

We are thus brought to the subject of altars. These are sufficiently varied in form. In the Kouyundjik bas-relief (Fig. 42) we find those shapes at the four angles which were copied by the peoples of the Mediterranean, and led to the expression, "the horns of the altar." In the Khorsabad relief (Fig. 107) the salience of these horns is less marked. On the other hand, the die or dado below them is fluted. Another altar brought from Khorsabad to the Louvre is quite different in shape (Fig. 108). It is triangular on plan. Above a plinth with a gentle salience rises the altar itself, supported at each angle by the paw of a lion. The table is circular, and decorated round the edge with cuneiform characters.

A third type is to be found in an altar from Nimroud, now in the British Museum (Fig. 109); it dates from the reign of Rammanu-nirari, who appears to have lived in the first half of the eighth century before our era.[321] The rolls at each end of this altar are very curious and seem to be the prototype of a form with which the Græco-Roman sarcophagi have made us familiar.

Fig. 110.—Stele from Khorsabad. Plan and elevation; from Place. Fig. 110.—Stele from Khorsabad. Plan and elevation; from Place. The various kinds of steles are also very interesting. The most remarkable of all is one discovered at Khorsabad by M. Place (Fig. 110). The shaft is composed of a series of perpendicular bands alternately flat and concave, exactly similar to the flutes of the Ionic order. The summit is crowned by a plume of palm leaves rising from a double scroll, like two consoles placed horizontally and head to head. The grace and slenderness of this stele are in strong contrast to the usually short and heavy forms affected by the Assyrian architects, especially when they worked in stone. It is difficult to say what its destination may have been. It was discovered lying in the centre of an outer court surrounded by offices and other subordinate buildings; it has neither figure nor inscription.[322] The base was quite rough and shapeless, and must have been sunk into the soil of the court, so that the flutes began at the level of the pavement. M. Place suggests that it may have been a milliarium, from which all the roads of the empire were measured. We do not know that there is a single fact to support such an unnecessary guess.

The stele of which we have been speaking is unique, but of another peculiarly Assyrian type there is no lack of examples, namely, of that to which the name obelisk has, with some want of discrimination, been applied. The Assyrian monoliths so styled are much shorter in their proportions than the lofty "needles" of Egypt, while their summits, instead of ending in a sharp pyramidion, are "stepped" and crowned with a narrow plateau. (Fig. 111). These monoliths were never very imposing in size, the tallest is hardly more than ten feet high.

Whatever name we choose to give to these objects, there can be no doubt as to their purpose. They are commemorative monuments, upon which both writer and sculptor have been employed to celebrate the glory of the sovereign. A long inscription covers the base of the shaft, while the upper part of each face is divided into five pictures, the narrow bands between them bearing short legends descriptive of the scenes represented. It was, of course, important that such figured panegyrics should be afforded the best possible chance of immortality; and we find that most of these obelisks are composed of the hardest rocks. Of the four examples in the British Museum, three are of basalt and one only of limestone.

Another type of stele in frequent employment was that with an arched top and inclosing an image of the king. It is often represented on the bas-reliefs[324] (Fig. 42), and not a few examples of it are in our museums. When we come to speak of Assyrian sculpture we shall have to reproduce some of them. We find a motive of the same kind, but more ornate and complicated, in the bas-relief from Kouyundjik figured above (Fig. 112). A hunting scene is carved on a wall of rock at the top of a hill. A lion attacks the king's chariot from behind; the king is about to pierce his head with an arrow while the charioteer leans over the horses and seems to moderate the determination with which they fly.[325] The sculpture is surrounded by a frame arched at the top and inclosed by an architrave with battlemented cornice. The whole forms a happily conceived little monument; it is probable that it was originally accompanied by an explanatory inscription.

This analysis of what we have called secondary forms has shown how great was the loss of the Chaldæan architect and of his too docile Assyrian pupil, in being deprived—by circumstances on the one hand and want of inclination on the other—of such a material as stone. Without it they could make use of none of those variations of plan and other contrivances of the same kind by which the skilful architect suggests the internal arrangement of his structures on their façades. For such purposes he had to turn to those constituents of his art to which we shall devote our next section.

NOTES