dimensions, generally occupying a space of about twenty-four feet from front to back, and varying somewhat in width. The hut next the rampart at the east end of the row was, however, as much as thirty-five feet in length, and projected towards the north beyond the line of its companions. This suggests that a verandah may have run along the front, as was the case at Gellygaer and elsewhere, but no trace was discovered of posts for its support. The exceptional length of the row, which has been already noted (205 feet), was doubtless intended to compensate for its reduced width.

The soil on which these huts lay showed many traces of occupation. Pottery was of frequent occurrence, especially on the site of Block No.11, where a fine bowl of Lezoux ware, bearing the stamp of Cinnamus,[1] and another decorated bowl,[2] as well as some coarser dishes, were discovered. At the west end of Block No. IV a number of iron objects—portions of wheel tyres and rings from axles—were found. Many small bronze objects, enamelled trinkets, and coins also came from this area. The most interesting find was, however, a bronze oenochoe,[3] which was taken out of a rubbish pit under Block XII. The pit was clearly of earlier date than the building, for the dividing walls of the two huts at the east end were built over it, and its existence was only ascertained by observing a slight subsidence of the foundations above it. The pit was not the sole evidence of an earlier occupation of the Praetentura. To the south the ends of the rows lay over the ditches of the Agricolan fort, and between Blocks I and III there were discovered the foundations of a large building, which, since it crossed the Agricolan ditch, could not have belonged to the first occupation, but must yet have been earlier than the lines of huts above it. Trenches had already been driven across the Praetentura from north to south, in hopes of finding earlier buildings, but without result, when in February, 1908, in cutting across the line of the Agricolan ditch, the workmen came upon the heavy foundations of the building of which we have been speaking. They lay at a considerable depth a trench had been cut down some two feet into the subsoil, and in this had been laid heavy sandstone blocks, from three feet six inches to three feet nine inches thick. As illustrating the level of the building, it may be noted that the top of the foundation at the east end lay from seven feet ten inches to five feet three inches below the modern surface, while in the case of the later buildings, the distance from the surface to the bottom of the foundations was only three

1 Plate xliv.

2 Plate xl. fig. 14.

3 Plate lv.