as at Housesteads and elsewhere, the corridors were built up to form rooms. It is possible that these pillars did not rise to any great height.

On the S. court, 2 ft. from the entrance, a small semi-circular base, 1 ft. 9 in. long and 2 ft. broad, was found. It had large foundations, one layer of pillarstones lying at the base. A small piece of moulding, 2 in. broad, might have been the border of an inscription, and the erection was possibly a stand for a statue.

The finds on the S. side comprise a few fragments of pottery, the top of a jar (amphora), some quoit-like stones (possibly baluster stones or weights), a quantity of lead, which had evidently been in contact with fire, and a few fragments of moulded slabs. Two of these were found in loose earth, and look like wall linings. The largest piece is so broad, and the mouldings so shallow, that it could not have been the border of an inscription.

The finds on the N. side were fewer, and comprise ordinary pottery, and an iron knife 6 in. long.

In this court were found very little signs of wood, and very few nails, but these appeared elsewhere in large quantities; they varied from 6 in. to 2 in., with a few of 1 in. made with big heads.

The N.E. corridor was floored with 1 ft. of gravel, and here, as in the other corridors, were a large number of flue-tiles, and one post-hole, with one single stone, 2 ft. below ground level, 7 ft. from the N. wall.

The most important find was an inscribed stone.

On the top line the letters P'. P'. were very clearly cut in letters 3 in. long. Then came the portion of a letter, probably a c, below was the letter A, and before this was a much mutilated letter, possibly a or M. All the letters were beautifully sculptured.

The N. corridor had a floor of 1 ft. 6 in. of stamped clay. There were the remains of at least two postholes, possibly three. These post-holes, both in the N. and S. courts, are unlike the other post-holes found in the Prætorium, the latter were simply holes driven

into the original clay, with a large stone at the bottom. But the former consist of a stone sunk 6 in. or 8 in. into the stamped clay about 1 ft. square, surrounded with a number of stones set up on end to keep the butt of the post in position. They may have been placed to give an additional support to the roof, or all these post-holes may have formed part of an earlier wooden Prætorium.

The support at the W. end is of rather a different character, consisting of one upright stone, with two stones set on end, and one stone at the base, possibly a door-catch.

The floor of the N.W. corridor was 3 in. of gravel over 10 in. of stamped clay. At the S. end was a post-hole, which may have been used as a door post.

The floor of the S. corridor was composed of 15 in. of clay and gravel, mixed. A row of post-holes corresponded to those on the opposite side, except that in this case the builders had made use of huge boulders, whose ends stand 8 ft. out of the natural ground. A drain, already described, runs down the N. side.

The finds were:-a small bronze coin (P. Antoninus 1) and a small amount of pottery.

A small post-hole 11 ft. E. of buttress near the outer wall, between this and the buttress a large amount of charcoal or decomposed wood under the clay.

The floor of S.E. corridor was 1 ft. gravel and 4 in. clay. At 8 ft. from door a large amount of charcoal. The inner court on the E. side was bounded by two

large squares of stone masonry, both originally, I should say, 6 ft. square. In a direct line with these was a buttress, 4 ft. 6 in. long by 3 ft. 6 in., having foundations 4 ft. deep of river boulder and dressed stones placed on their edges, many of them from 3 ft. to 4 ft.

long, and capable of bearing enormous weight. The wall of the Sacellum protrudes here 2 ft., and the wall of the smaller chamber, contrary to all the other interior walls, is 2 ft. thick. From this I would suggest that there was an upper storey, and, for reasons given further on, I think that it was carried over the two chambers N. of the Sacellum. These conclusions may be altered when the remaining rooms have been excavated.

The floor was 10 in. of stamped clay.

By the side of the chamber (9) was a small brick platform, 4 ft. long, 2 ft. wide, and 1 ft. high. In front of the platform were two post-holes.

Two more post-holes occur between the two squares of masonry. Two occur between the Sacellum and the most southern square of masonry, about 1 ft. square. By the buttress were found a number of bricks, and a quantity of broken pavement.

Finds.-A large quantity of nails, two fibulæ, broken pottery, and an iron instrument, possibly a bradawl.

The roadway up to the centre had a gravel floor 12 in. deep.

The inner court on the S. side is bounded by a 2-ft. wall with deep foundations, which is carried on for a distance of 10 ft. In line with this is a square of masonry, 6 ft. by 4 ft., similar to those on the opposite side, and by the outer wall of the Prætorium there is a buttress.

The floor on the N. side is 1 ft. clay, on the S. side 2 ft. At the E. corner of the wall is a post-hole sunk in the solid clay, and between the square and buttress we found a support made of stones to take a post.

By the outer wall of the Prætorium is a thick square of masonry, 2 ft. square, which may possibly be the site of a stone pillar, as the workmen told me (this portion was opened before I arrived) that a number of pillarstones had been found here. The floor was paved.

Finds.-Several pieces of pottery, a silver coin (Antoninus), and a bronze coin. By the side of the buttress I found a door-hanger.

CHAMBERS OPENING FROM INNER COURT.

At the N. end is a room (9) in an unusual position. The floor was 8 ft. gravel, and the inside wall between it and the inner court had brick foundations.

There were no finds of importance here.

The angle chamber W. of Room 10 measures 10 ft. 6 in. by 13 ft., beyond which is a larger room. Both these chambers had their floors higher than the remainder of the Prætorium. Only a portion of the N. wall remained. The floor was composed of 2 in. of brick cement (formed of lime and pounded brick), 3 in. clay, 4 in. brick-cement, and 12 in. of clay. On the second floor of brick-cement (4 in.) was found a large amount of small, selected water-worn stones, possibly slingstones. I do not consider that their occurrence here is accidental. There were four post-holes in Room 10, but they may have belonged to an earlier building, later excavations may confirm this hypothesis. Possibly