5.02 m. high, so that the structure as a whole must have been an almost perfect cube. . In the middle of the ground-beam on the western side there was an indentation 1.83 m. broad, where the door had been; the door, which was 2.3 m. high, had faced N. 57° W. The two side-posts for the door were pretty strong, and were fixed into the bottom beam in the manner shown in the accompanying cut (fig. 281). The two lintels consisted each of a short plank fastened to vertical side-planks, and each half-door swung upon a pivot in the ground-beam and upon another pivot in the lintel. The ground-beams were 33 cm. broad and 2I cm. high. Considering the fragmentary condition of the ruins, it was impossible to attempt a reconstruction of the house.
Fig. 279. OUR CAMP OF MARCH 28.
The other house, or perhaps rather row of houses, was of greater dimensions. It lay 20.4 m. west-north-west of the first one, and, like it, stood upon a platform 21/2 m. high. The longer side of the entire complex measured 52.4 m., and was situated N. 22° E. to S. 22° W., that is almost parallel to the longer side of the first house. The windward sides of both mounds were seriously damaged by the wind; in the case of the larger house the tie-beams of the foundations were entirely exposed at their east-south-east ends and hung in the air a couple of meters above the ground. Between the two mounds there was a hollow or passage, which continued a good way farther in the direction S. 4o° W., though there it was broader, and was inclosed between terrace-like jardangs. The fact of the two long walls of both houses lying parallel to the direction of the prevailing wind would seem to indicate, that the inhabitants had endeavoured to protect themselves as far as possible against it, though it would, it is true, be less felt when the country was far and wide planted with forest. Their position has however facilitated the erosive and corrasive action of the wind upon the dry clay soil, and this is the cause of the distinctly marked hollow between the two houses. It is however rather due to accident that the south-westward prolongation of the nearest gully forms an immediate continuation of the hollow; the former would no doubt have come equally into being had the houses and their mounds never existed.
A closer examination of the beams I have lately alluded to as hanging suspended in mid-air proved beyond a doubt, that they were originally laid upon per-
IIIF Color HighRes
