CELTIC AND ROMANO-BRITISH HERTFORDSHIRE 



drapery indicated that there were also figures. The entrance to this 

 chamber must have been from the north-east side, where the wall has been 

 destroyed to a foot below the floor level, and from its present appearance 

 seems to have been a sleeper wall for a continuous course of blocks of stone 

 upon which was possibly a colonnade. On account of the depth of soil Uttle 

 of the floor of tessellated pavement could be examined. From what was 

 exposed, however, it was ascertained that it had an outer border of coarse 

 drab tesserae (the tesserae being about i in. by liin.) which extended from 

 the side walls about 5 ft. 6 in. and rather more from the end walls. Within 

 there was a border of a scale pattern in smaller black and white tesserae, and 

 within this again was a very pretty wide braid-work design in black, red, 

 drab and white, and then lines of black and white. Within these borders 

 was the main part of the pavement, which was much mutilated owing to the 

 fallen masonry having become embedded in it, but it appeared that the design 

 was geometric, made up of a series of bands of a scroll pattern in very small 

 red, white, yellow and black tesserae. It is difficult to suggest a purpose for 

 this chamber. It may, perhaps, have been a court connected with the forum, 

 the raised apse forming the tribunal. 



South-east of this chamber was a narrow passage 3 ft. 2 in. in width, 

 at the bottom of which about 6 ft. or 7 ft. down was a deposit of black 

 mud ; but as each end of the passage was blocked by very solid walls it is 

 difficult to understand how it can have been a drain or waterway. Beyond 

 this is a coarse red tessellated pavement 6 ft. 6 in. wide and 76 ft. long, 

 bounded on its north-western side by a roughly built wall 2 ft. 6 in. at the 

 footings and i ft. 6 in. above and on its south-eastern side by a sleeper wall 

 2 ft. 8 in. wide. Beyond this again is another pavement of coarse red 

 tesserae, in a good state of preservation, 8 in. above the level of the last- 

 named pavement. It is 21 ft. wide and 76 ft. long, and is bounded on its 

 south-east side by the sleeper wall adjoining the street. 



The fact that the inner wall of the south-west corridor, which has been 

 fully excavated, is in its north-western half practically an exact duplicate 

 of its south-eastern half, with foundations for pilaster buttresses in the same 

 position, suggests a duplication of the plan. 



At an uncertain date, possibly in the latter part of the Roman occupa- 

 tion, the whole or nearly the whole of the corridors, at all events those on 

 the south-west and north-east sides, were burnt and ruined. Over the 

 corridors on these sides was a layer of charcoal, and at one spot was a 

 considerable quantity of molten lead, which must have melted at a high 

 temperature and fallen in a molten state from a considerable height,'^ while 

 numerous flints and bricks showed evidence of the action of fire. The 

 building must have remained in a half-ruinous state for some time, for over 

 the layer of charcoal is an accumulation of rubbish a foot deep, on the top or 

 which a fresh floor level was made and the corridors were patched and altered. 

 At this time apparently the south-east and north-west corridors were cut off 

 by low walls i ft. 2 in. above the old level, possibly to allow for the change 

 of level of the south-west corridor. That on the south-east side has two 

 channels paved with bricks, which, on comparison with similar work at 



" Information from Professor Gowland, who kindly analysed some of the lead. 



