but the robbed backwall had rested only on the 

 natural sandy clay at a depth of 2 feet 3 inches below 

 the modern grade. In front of the chimney, and rising 

 from the cellar floor, was a massive brick-walled 

 underhearth 7 feet 6 inches wide and projecting out 

 from the fireplace to a distance of 5 feet. 



A curious and still unexplained feature of the under- 

 hearth was a 4- by 3-inch channel running across the 

 top of the surviving foundation for a distance of 

 6 feet 9 inches, starting at the south face and terminat- 

 ing 9 inches short of the north. This channel had 

 been bricked over and the remaining bricks had 

 dropped into it (fig. 5) presumably after a wooden 

 beam, which once occupied the space, had rotted or 

 burned out. Traces of burned or carbonized wood 

 lay on the clay bottom of the channel, but the bricks 

 over it displayed no evidence of fire. The only con- 

 ceivable explanation for the presence of the wood 

 must be that it was part of a frame used to hold the 

 block of natural sandy clay together while the under- 

 hearth wall was being erected around it. As the 

 underhearth foundation would have originally risen 

 at least another 2 feet 6 inches above the timber to 

 the floor level of the house, the wood would not have 

 been in danger of igniting from the heat of the domes- 

 tic fire. But if the house ultimately burned, it is 

 possible that the exposed end of the timber might 

 have caught fire and slowly been consumed along its 

 entire length. 



The cellar hole had been cut into natural sands 

 clay to an average depth of 5 feet 3 inches below the 

 modern grade. Its backfilling was predominantly of 

 the same sandy clay and, consequently, the exact 

 edge of the cellar hole was sometimes hard to deter- 

 mine. It was probably because of this similarity 

 between the natural subsoil and the cellar's fill that 

 the feature failed to show up in the soil resistivity 

 survey. Owing to previously mentioned limiting 

 factors, only the southeast corner of the cellar hole 

 was found and only parts of the south and east walls 

 were traced out. Consequently, it can merely be said 

 that the cellar exceeded 27 feet in east west length 

 and 11 feet 2 inches in width (fig. 3). 



Three post holes were found against the south face, 

 while the rotted remains of another vertical post were 

 found north of the chimney supporting a much- 

 decayed horizontal board that had served to revet 

 the east face. A broad-bladed chisel (fig. 14, no. 6) 

 was found behind the board where it had probably 

 been lost while the timbering was being installed. 



Further slight traces of horizontal boards were found 



along the south face, suggesting that the soft sides of 

 the large cellar hole had been supported in this way. 

 But it was not possible to determine whether the 

 boards had been placed only on sections of the wall 

 that seemed in danger of sliding in or whether the 

 entire interior had been sheathed with planks. The 

 south side of the cellar hole sloped outwards at an 

 approximate 65 percent angle and the traces of boards 

 lay against it. 1 ' However, it was not possible to tell 

 whether the vertical posts had been similarly sloped, 

 but it is reasonable to assume that they would have 

 done so. 



Parts of the cellar's wooden floor still survived (figs. 

 6 and 7) and comprised boards ranging in width from 

 5 to 7 inches laid over sleepers or joists 4 to 6 inches 

 wide. The height of the underlying timbers could 

 not be determined as the weight of the cellar fill might 

 be assumed to have pressed the floorboards down as 

 the wood of the sleepers decayed. Only occasional 

 floorboards survived and the channels left by decayed 

 sleepers did not extend across the full width of the 

 excavated cellar. From these facts it was deduced 

 that the boards had been cut from woods of different 

 types, some of which had decayed more completely 



15 Albert C. Manucv, ■'The Fort at Frederica," Notes in 

 Anthropology (Tallahassee: Florida State University, 1962), vol. 

 5, pp. 51-53. An excavated powder magazine of 1736 exhib- 

 ited similar construction. 



Figure 5. — Detail of collapsed bricks in the under- 

 hearth. (Pfwto courtesj oj E DeHardit.) 



PAPER 52: EXCAVATIONS AT CLAY BANK 



