CONSTRUCTION. 189 freshness. It seems to be proved that in at least the majority of rooms at Khorsabad the architect provided other means of lighting and ventilation besides the doorways, wide and high though the latter were. He pierced the roof with numerous oblique and vertical openings, he left square wells in the timber ceilings, and circular eyes in the domes and vaults. If these were to fulfil their purpose of admitting light and air into the principal rooms, the latter must have had no upper stories to carry. At Mossotil, walls are much thinner than at Nineveh, and interiors are simpler in arrangement and decoration. The twenty or five-and-twenty feet of clay of the Assyrian walls would make it impossible to give sufficient light through the doors alone to the sculptures and paintings with which the rooms were adorned. Ve cannot doubt that a top light was also required. The rooms of the palaces must, therefore, have suc- ceeded one another in one horizontal plan. Slight differences of level between them were connected by short flights, usually of five carefully-adjusted steps. 1 In spite of all its magnificence the royal dwelling was no more than a huge ground floor. With such methods of construction as those we have described, it would have been very difficult to multiply stories. Neither vaults nor timber ceilings could have carried the enormous masses of earth of which even their partition- walls for the most part consisted, so that the architect would have had no choice but to make his upper chambers identical in size with those of his ground floors. This difficulty he was not, however, called upon to face, because the necessity for providing his halls and corridors with a top light, put an upper floor out of the question. No trace of such a staircase as would have been required to give access to an upper story has been discovered in any of the Assyrian ruins, 2 and yet some means of ascent to the terraced roofs must have been provided, if not for the inhabitants of the chambers below who are likely, however, to have passed the nights upon them in the hot season at least for the workmen whose duty it was to keep them in repair. Some parts of the palace, on the other hand, may have been raised much above the level of the rest. Sir Henry Layarcl found the remains of such chambers in the palace of Assurnazirpal at 1 PLACK, Ninire, vol. i. p. 307. - See BOTTA. Monument de JVinive, vol. v. p. 53 ; PLACE, Ninire, vol.1 pp. 306, 307,
