When the southern extension of this deposit was excavated in 2017, it was interpretated as a collapse context (17040 = 17095). The deposit may have been over-excavated into the charcoal-rich destruction deposit that it covers. We can possibly also equalize 20099 = 17040 = 17095 with 17053 and 17031, both east of wall 17012 and characterized by a dense accumulations of small shattered limestones in an orange-brown matrix over a charcoal-rich context (17033 and 17050). In 2022 we have reinterpreted this deposit of tightly packed tile and stone as a deliberate levelling layer associated with the re-occupation of Room D after the mid-1st century fire. it is possible that the new occupants reused materials associated with the destruction event to level out the room and create a preparation layer on top of which they laid a pavement. The large number of structural nails found in these deposits may suggest, however, that the deposit is the collapsed superstructure of the walls. The stone and tile fragments may have been used as a sort of aggregate, packed into clay to form pise or rammed earth walls, which collapsed during/after the mid-1st c. CE fire. This wall material could have been levelled out by later occupants and paved over.

Collapse layer used as preparation for later floor levels - EBS