Fig. 41.—Temple on the bank of a river, Khorsabad; from Botta.
We cannot say the same of the arch, which played a much more important rôle than it did in Egypt. There it was banished, as we have seen, to the secondary parts of an edifice. It hardly entered into the composition of the nobler class of buildings; it was used mainly in store-rooms built near the temples, in the gateways through the outer walls of tombs, and in underground cellars and passages.[162] In Mesopotamia, on the other hand, the arch is one of the real constituent elements of the national architecture.
Fig. 42.—Temple in a Royal Park, Kouyundjik; from the British Museum.
That the Chaldæan architects were early led to the invention of the arch is easily understood. They were unable to support the upper parts of their walls, their ceilings or their roofs, upon beams of stone or timber, and they had to devise some other means of arriving at the desired result. This means was not matured all at once. With most peoples the first stage consisted probably in those corbels or off-sets by which the width of the space to be covered was reduced course by course, till a junction was effected at the top; and sometimes this early stage may have been dispensed with. In some cases, the workman who had to cover a narrow void with small units of construction may, in trying them in various positions and combinations, have hit upon the real principle of the arch. This principle must everywhere have been discovered more or less accidentally; in one place the accident may have come sooner than in another, and here it may have been turned to more profit than there. We shall have to describe and explain these differences at each stage of our journey through the art history of antiquity, but we may at once state the general law that our studies and comparisons will bring to light. The arch was soonest discovered and most invariably employed by those builders who found themselves condemned, by the geological formation of their country, to the employment of the smallest units.
The Chaldæans were among those builders, and they made frequent use of the arch. They built no long arcades with piers or columns for supports, like those of the Romans, and that simply because such structures would have been contrary to the general principles of their architecture. They made no use, as we have already explained, of those isolated supports whose employment resulted in the hypostyle halls of Egypt and Persia, in the naves of Greek temples and Latin basilicas. The want of stone put any such arrangement out of the question. We have, then, no reason to believe that their arches ever rested upon piers or upon the solid parts of walls freely pierced for the admission of light. The type from which the modern east has evolved so many fine mosques and churches was unknown in Chaldæa. In every building of which we possess either the remains or the figured representation the archivolts rest upon thick and solid walls.
Under these conditions the vault was supreme in certain parts of the building. Its use was there so constant as to have almost the character of an unvarying law. Every palace was pierced in its substructure by drains that carried the rain water and the general waste from the large population by which it was inhabited down into the neighbouring river, and nearly all these drains were vaulted. And it must not be supposed that the architect deliberately hid his vaults and arches, or that he only used them in those parts of his buildings where they were concealed and lost in their surroundings; they occur, also, upon the most careful and elaborate façades. The gates of cities, of palaces and temples, of most buildings, in fact, that have any monumental character, are crowned by an arch, the curve of which is accentuated by a brilliantly coloured soffit. This arch is continued as a barrel vault for the whole length of the passage leading into the interior, and these passages are sometimes very long. Vaults would also, in all probability, have been found over those narrow chambers that are so numerous in Assyrian palaces were it not for the universal ruin that has overtaken their superstructures. Finally, certain square rooms have been discovered which must have been covered with vaults in the shape of more or less flattened domes.