During the HCP work for the regulation of rain and spring water in the southern corner of the ancient beach of Herculaneum, excavations were undertaken on an area of the beach that had not been previously investigated. In the area between the southern corner of the Suburban Baths and the south wing of the House of the Telephus Relief. Around one metre of solidified mud from the 79 A.D. eruption was removed. This revealed an imposing collapse of timbers which, thanks to the conditions of their burial, were perfectly preserved and not carbonized. The position of the fall indicated that they belonged to the primary and secondary vertical beams from the roof of the so-called “Hall of the Marbles” (room 18) in the House of the Telephus Relief. The investigations showed that the roof seemed to have overturned due to the “sucker effect” caused by the first mud flow which ripped it from the tops of the walls, overturning it and hurling it onto the beach. In fact, numerous tile fragments from the roof and tufelli from the upper parts of the walls, which had been torn off together with the timbers, were found below the beams in direct contact with the sand of the ancient beach.

Large beams with a rectangular section and smaller beams with square or circular sections emerged from the blanket of mud surrounding them, together with other timber elements including planks and decorated panels, the latter probably part of the room’s false ceiling. Marks left by the carpenters were clearly visible, as were the joints, on all the elements. On the basis of a first analysis of the beams, a study which will produce important information regarding Roman carpentry, it may be suggested that this was a pitched roof with trusses. The recovery of planks painted in red or blue and of diverse wooden panels with parts of a wooden coffering and frames with hexagons and triangles in relief, painted white, black, blue, red and gold, revealed the presence of a richly decorated false ceiling which could find correspondence in the pattern of the room’s complex marble floor.

After having brought all the elements to light they were numbered and a 3D laser scan was made of the entire collapse, so as to have a three dimensional model of the finds. The individual elements were then removed, noting all relationships which could be of use in the study phase for the reconstruction of the roof. Each find was then photographed, measured and the joints and any traces of colour recorded where present. With the assistance of the conservators from the Pragma cooperative each find was then cleaned and treated with biocide. They were then individually wrapped and identified with a number assigned at the time of their discovery.