70 AN EAULY NEOLITHIC KITCHEN-MIDDEN. 



In various parts of the pit miniature ramifying caves may IDC 

 seen, corresponding exactly, on a smaller scale, with the caverns 

 found in an ordinary limestone. These caves usually contain a 

 whitish deposit of calcareous cave-earth sealed up beneath a hard 

 thin stalagmitic crust. Above the crust is sometimes another 

 blacker deposit containing shells like those of the Roman layer or 

 soil, though below the crust one only finds the Neolithic species. 



At the southern end of the pit Mr. Wallace and I discovered 

 still further evidence of the antiquity of the deposit, for a grave 

 had been sunk about four feet into the tufa, lined with slabs of 

 Purbeck stone, and contained the skeleton of a youth buried in 

 a contracted position. This was apparently an interment consider- 

 ably older than the Roman period, probably Neolithic. Mr. 

 Wallace, who examined the grave, could find nothing in it besides 

 the skeleton, but it had already been broken into before our visit. 

 No deposit of tufa had taken place after this grave was dug. The 

 Roman layer apparently passed over the grave, though that point 

 was not perfectly clear. Thus we can prove that the tufa is not 

 only earlier than the Roman layer, but had ceased to form before 

 this interment took place. 



The next point to decide is : how much earlier is the tufa than 

 the Roman layer, and for this purpose we have only the fossil 

 contents to guide us, for it rests directly on Wealden Beds. The 

 contents of the tufa are so singular as at first to make me think 

 that the deposit might date as far back as the Palaeolithic Period ; 

 but closer examination convinces me that it cannot well be older 

 than early Neolithic. Taking first the ethnological evidence : we 

 have not yet discovered human remains, though we have abundant 

 evidence that man lived in the neighbourhood during the whole of 

 the period when tufa was being deposited. Flint-flakes and 

 charcoal occur throughout, though they are most abundant about 

 the middle of the deposit. With them we find cores from which 

 flakes have been struck, and occasionally a rough chalk-flint 

 apparently thrown away as worthless. The flaking is of the 

 ordinary Neolithic type, though poorly done and the material badly 



