Fig. 61.—Armenian "Lantern;" from Botta.
"The houses of the Armenian peasantry," he says, "are sunk into the ground, so that their walls stand up but little above the level of the soil. They are lighted by an opening that serves at once for window and chimney, and is placed, as a rule, in the centre of the roof. The timber frame of this opening is often ingeniously arranged ([Fig. 61]). Four thick beams, but very roughly squared, intersect each other in the middle of the house. Across their angles slighter joists are placed, and this operation is repeated till a small dome, open at the top, for the entrance of light and the escape of smoke, has been erected."[225]
Figs. 62-65.—Terra-cotta cylinders in elevation, section and plan; from Place.
In the case of vaults how are we to suppose that the rooms were lighted? We can hardly imagine that rectangular openings were left in the crown of the arch, such a contrivance would have admitted very little light, while it would have seriously compromised the safety of the structure. According to M. Place the desired result was obtained in more skilful fashion. In several rooms he found terra-cotta cylinders similar to those figured below. These objects, of which he gives a careful description, were about thirteen and a half inches in diameter and ten inches in height. We may refer our readers to the pages of M. Place for a detailed account of the observations by which he was led to conclude that these cylinders were not stored, as if in a warehouse, in the rooms where they were afterwards found, but that they formed an integral part of the roof and shared its ruin. We may say that the evidence he brings forward seems to fairly justify his hypothesis.
Penetrating the roof at various points these cylinders would afford a passage for the outer air to the heated chamber within, while a certain quantity of light would be admitted at the same time. The danger arising from the rains could be avoided to a great extent by giving them a slightly oblique direction. To this very day the Turkish bath-houses over the whole of the Levant from Belgrade to Teheran, are almost universally lighted by these small circular openings, which are pierced in great numbers through the low domes, and closed with immovable glasses. Besides which we can point to similar arrangements in houses placed both by their date and character, far nearer to those of Assyria. The Sassanide monuments bear witness that many centuries after the destruction of Nineveh the custom of placing cylinders of terra-cotta in vaults was still practised. In spite of its small scale these circles may be distinguished in the woodcut of the Sarbistan palace which we have borrowed from Coste and Flandin ([Fig. 54]).[226]
These same writers have ascertained that the architects of Chosroes and Noushirwan employed still another method of lighting the rooms over which they built their domes. They gave the latter what is called an "eye," about three feet in diameter, through which the daylight could fall vertically into the room beneath. This is the principle upon which the Pantheon of Agrippa is lighted; the only difference being one of proportion. In Persia, the diameter of the eye was always very small compared to that of the dome. If we are justified in our belief that the constructors of the Parthian and Sassanide palaces were no more than the perpetuators of systems invented by the architects of Nineveh and Babylon, the Assyrian domes also may very well have been opened at the summit in this fashion. In the bas-relief reproduced in our [Fig. 42], the two small cupolas are surmounted with caps around a circular opening which must have admitted the light. Moreover, the elaborate system of drainage with which the substructure of an Assyrian palace was honeycombed would allow any rain water to run off as fast as such a hole would admit it.[227]
Whatever may be thought of these conjectures, it is certain that the architects of Nineveh—while they did not neglect accessory sources of illumination—counted chiefly upon the doors to give their buildings a sufficient supply of light and air. As M. Place says, when we examine the plans of Sargon's palace at Khorsabad we are as much astonished at the size of the doorways as at the thickness of the walls.[228]