Another very early arch was discovered by M. De Sarzec at Tellô, close to the building of Ur-Ninâ (cf. Fig. [16]), having much the same shape as the Nippur arch illustrated in Fig. [15] and doubtless used for a similar purpose, while vaulted passages of which the arch was semicircular, were discovered by Taylor[84] in his excavations at Muḳeyyer (Ur), as early as 1855.

Again the German excavations at Fâra (Shuruppak) in 1902 and 1903 revealed a number of circular rooms, each of which was roofed by means of an arch formed by overlapping bricks placed horizontally, somewhat after the fashion of the later corbelled arch at Nippur seen in Fig. [17], to which Hilprecht assigns a provisional date of 2500 B.C. We know that the dome was invented in Babylonia at a very early date, thanks to Dr. Banks’ discovery at Bismâya of an oval-shaped room in the vicinity of the temple, the lower parts of the domed roof of which were found still in place. Its antiquity is attested by the date of the temple itself which would appear to have belonged to the pre-Sargonic period, as the ziggurat was faced with the plano-convex bricks characteristic of that period, and the pottery furnace, not far distant, was composed of bricks of the same kind.

In later times the arch was doubtless used more frequently in Babylonia. A good example of a late Babylonian arch was discovered by the German excavators on the Ḳasr at Babylon; the arch in question (cf. Fig. [18]) which is Roman in character, forms the roofing of a lofty gate cut in the fortification wall. Koldewey[85] is of opinion that the wall in which this arched gate occurs, is a good deal older than the Nebuchadnezzar period.

But Assyria, the more or less faithful imitator of Babylonia in all matters great or small, is also known to have employed the arch as an architectural device, though, as in Babylonia, most of the Assyrian arches which the excavations have brought to light are connected with the drainage system with which all the principal buildings were provided. The best examples of an Assyrian arch of ordinary dimensions are those found at Khorsabad, the gateways of which town were roofed with semicircular vaults. One of these gateways was pulled down by Place in order to make a close examination of its construction. The height from the pavement to the top of the arch was found to be twenty-four and a half feet, the width being a little over fourteen feet. The arch was made of crude bricks, all of which were of the same size, and had the same shape, the bricks being cemented together with soft clay. The vault itself had long since become disintegrated, but the materials of which it was made were discovered in the ruins. Of the brilliantly painted friezes which adorned these rounded openings (cf. Pl. [XXX]), something will be said in the chapter on Painting.

Fig. 17.—Corbelled Arch at Nippur.
(Cf. Hilprecht, Explorations, p. 420.)		Fig. 18.—Arch at Babylon.
(Cf. Mitteil., 8, Abb. 1.)

But in regard to the study of what may be called the arch-principle, the subterranean channels which formed part of the system of drainage employed by the Assyrians are of greater importance. These aqueducts are found in all the palaces, both at Nimrûd and Kouyunjik, but Khorsabad furnished the best preserved examples, and therefore afforded the most valuable material for the careful examination of this architectural contrivance. At Khorsabad, Place discovered several arched drains of different shapes, some of them being round, others elliptical, while others again were pointed, but apparently in every case the stones or bricks were set at an angle, so that each course had the support of the course preceding it, and thus the pressure on the centre of the arch was reduced to the minimum. In the case of the pointed arched aqueducts found by Place, the arches in question are no true keystone arches, indeed they have no keystones of any kind as will be seen in Figs. [19], [20]; this arched drain measures four feet eight inches from the ground to the centre of the vault, its width is about three feet nine inches, while its original length is unfortunately not known, though Place succeeded in tracing it for some two hundred and twenty feet. The floor was made of large slabs of limestone set in asphalt, while the ends of the stone-slabs extended beyond the walls of the vault on either side. The rounded type of arch is seen in Fig. [21]; it is semicircular in shape and is formed of three voussoirs on each side, which together with the key, thus make seven in all, but owing to some miscalculation, the keystone appears to have been too small, in consequence of which there was a gap between it and the top voussoir on the right, which was filled in by means of a stone wedge which can be seen in the figure. Its width and height vary at different points, in some places it is said to be wide enough for two men to walk abreast in it.[86] The floor was composed of slabs of limestone which were laid in asphalt just like the floor of the pointed arch described above. An elliptical-shaped arch, also found at Khorsabad, is illustrated in Fig. [22]: it is formed of eight voussoirs and a keystone, the gap on either side of which is filled in by means of two stone wedges. The failure to make the keystones sufficiently large, and the consequent necessity for these supplementary wedges may be due to the architects not having allowed for the shrinkage of the bricks.

Fig. 19. (Place, Nineveh, Pl. 38.)Fig. 20. (Place, Nineveh, Pl. 38.)
Fig. 21. (Place, Nineveh, Pl. 39.)Fig. 22. (Place, Nineveh, Pl. 39.)

In regard to the arched structures at Nimrûd (Calah) Layard says he found a vaulted room and more than one arch. He tells us that “the arch was constructed upon the well-known principle of vaulted roofs, the bricks being placed sideways, one against the other, and having been probably sustained by a framework until the vault was completed.” Knowledge of the principle of dome-shaped roofs in Assyria as well as in the mother-country, is evidenced both by the discovery of rooms whose dimensions would have rendered any other mode of roofing impossible, and also by the representations on Assyrian bas-reliefs, as we have already seen.[87]

Fig. 23. (After Andrae.)Fig. 24. (After Taylor.)

The arch-principle is further embodied in some of the Babylonian and Assyrian graves, and as there is no other opportunity of discussing the burial-places of the Babylonians and Assyrians in this volume, it may be permissible to give here a brief and general description of one or two of the best preserved of these burial-vaults. At Muḳeyyer (Ur) Taylor found a number of arched vaults (cf. Fig. [24]) which in most cases measured about 5 feet in height, and 3 feet 7 inches in breadth, while they were about 7 feet long at the bottom and 5 feet long at the top. The arch is formed by successive layers of overlapping bricks. It is interesting to compare the burial vaults discovered by Andrae at Ḳalat Sherḳat (Ashur), one of the best preserved of which is seen in Fig. [23].[88] This vault was discovered about 16 feet below the level of the floor of a Parthian door in the neighbourhood, and over 13 feet below a later Assyrian pavement. At the time of its construction the vault would appear to have been about 9 feet beneath the surface of the soil. The perpendicular walls forming the sides of the lower part of the tomb are set upon a brick pavement, the bricks being about 10-1/2 inches square and nearly 2-1/2 inches thick. The height of these perpendicular walls is approximately 30-1/2 inches and the layers of bricks which each contains number 13. The vault itself, which of course commences where the perpendicular walls cease, is more or less oval in shape, has a span of 5 feet 2 inches, and is 2 feet 11 inches high, the total height of the tomb from the floor to the top of the arch thus being nearly 5-1/2 feet. The arch, upon the construction of which much care had evidently been bestowed, was formed of forty-six courses of quasi-wedge-shaped bricks, each resembling a truncated segment of a circle. The interstices between the courses were filled in with stones, broken pieces of clay and clay mortar. The outside of the vault was coated with clay, but the inside was left plain. The walls at either end incline inwards, while they are built separately from the arch and are also rather higher. Access to the tomb from outside is gained by a slanting and somewhat winding entrance-shaft built close to the western wall, in which there is a small arched opening. The threshold of this opening, which was filled in with loosely laid bricks, lies 23-1/2 inches above the floor of the tomb. In the eastern wall the usual recess was found, the floor of which was some 35 inches above the floor of the burial chamber. This recess was about 19 inches in height and 13 inches in breadth, while the depth of the recess was greater than that of the wall, and it was therefore necessary to build another wall on the outside to close it up. In the recess of another double grave found by Andrae on the same site, a clay lamp was found, but this was not the case here; possibly there was at one time a lamp in the recess of this tomb, its disappearance being due to the disintegration to which the long infiltration of damp subjects all articles of unburnt clay. The entrance-shaft was 39 inches square, while its bottom was level with the threshold of the small opening in the western wall; 4 feet 2 inches above this floor was a second floor made of gypsum blocks which were supported by the walls of the shaft. The interstices between these blocks were filled in with stones and pieces of clay, while the upper part of the shaft (i.e. the part above these blocks), the walls of which had only half the thickness of the walls below, was similarly filled in right up to the surface level. The uppermost part of the shaft had been disturbed by a later building. In the vault, Andrae found three skeletons, one of which apparently belonged to a man, and the other two to women. The arms of these skeletons were at right angles to the bodies, and the legs were contracted and apart, while the man lies on his right side and the two women on their left. Traces of a decayed whitish material were found in the tomb, which Andrae believes to be the remains of grave-clothes. Bone needles and pottery had also been deposited with the corpses. The most interesting articles of pottery were three wide-necked bottles, two of which were decorated with dark horizontal lines, while the neck of the third was adorned with white painting on a dark ground, a technique well known in early Assyrian times. What were the contents of the vessels we do not know; rams’ bones were found near the door as well as elsewhere, and without doubt these vessels once contained meat offerings and drink offerings for the dead. There was evidence for at least three different periods of occupation in the strata above the grave, two of which belonged to the Assyrian era and one to the Parthian times.