The outer circuit shared in this heightening to an equal extent, or, to speak more accurately, it was the continual heightening of the roads that lay around it that was the reason for raising the temple. The same arrangement can be seen to-day in Oriental cities. The newly-built houses are of course so constructed that the ground floor is on about the same level as the street. As the latter, however, serves as the depository of all sorts of rubbish it is not long before the ground floor is below the street level. In Bagdad, for example, one has always to step down on entering an old house from the street, and the older the house the deeper the step. When the building becomes ruinous and requires rebuilding, the new floor is of course made level with the street. Part of the rubbish of the destroyed house is used to raise the level of the house, the rest is thrown into the street. If the houses are built of burnt brick a large part of the building material can be re-used, but with houses of mud brick almost the whole of the material becomes rubbish, which when spread out gradually raises the whole area. It follows that in the course of hundreds or thousands of years such a town site must become very considerably higher (see Fig. 154). It must be taken into consideration that later and more cultured periods yield higher deposits of rubbish than earlier ones, which are remains of simpler conditions of life, and of unpretentious dwelling-places. Also in the course of a long period the rubbish is much more pressed together by its own weight than in a shorter period, when the process of compression has not been so prolonged. Thus in the 1700 years between Nebuchadnezzar and the eleventh century A.D., Amran rose 21 metres, while at Merkes, as we shall see presently, the mounds of rubbish, which are also the accumulation of 1700 years, from the time of Hammurabi 2250 B.C. to Nabonidus 550 B.C., rose only 6 metres. According to this we must reckon on a retrocessive sequence of the density of the layers, which is expressed in the figures 21 and 6. While in Amran we must reckon 80 years for every metre of depth of rubbish, in Merkes every metre represents 280 years. The application of even an approximately rapid sequence at Fara leads to a height of antiquity which at first we hesitate to accept, but to which we may have to accustom ourselves, as geology has accustomed itself to the remote periods which are now universally accepted for the genesis of certain strata. In spite of all these heightenings which were carried on in the temples, they rarely rose to any considerable height above their surroundings, and they were always on the same level as the city, in opposition to the highly placed temples at the zikurrats. Somewhat to the north of temple “Z” we made a transverse cut through the narrow back of the mound, and in the mud-brick houses that lay there we found a number of business and scientific tablets.