had. in fact, taken place there during the 1930s. My first seven years 

 of the new Swartkrans project were spent in supervising the removal 

 of miners' rubble from the cave's environs, sorting it and recovering 

 whatever fossils were to be found, although these obviously lacked 

 stratigraphic information. This interlude gave me the opportunity of 

 doing wide-ranging studies on cave taphonomy and on the charac- 

 teristics of carnivore food remains. Once clear of rubble and natural 

 hillside-soil overburden, it became apparent that we were dealing 

 with a cave deposit about 45 meters in both east-west and north- 

 south dimensions, with a vertical shaft at the northeast corner, that 

 led down into "the lower cave." This ramified beneath the north 

 wall of the fossil cavern and had connections upward to the latter. 

 To our surprise, we found that the heavily calcified breccia from 

 which Broom and Robinson had obtained nearly all their fossils 

 represented an isolated block attached to and hanging from the 

 cave's north wall. This became known as the "Hanging Remnant" 

 and it provided a useful clue to the understanding of the very com- 

 plex relationships of the deposits within the cave. As the excavation 

 proceeded. I searched for several years for the remnants of an earlier 

 deposit on which the Hanging Remnant sediment must have rested 

 before being undercut by erosion. Eventually we found this as a 

 talus cone of calcified cave earth that had accumulated below a long- 

 gone entrance above the cave's south wall. 



It transpired that the extraordinary complexity of the cave filling 

 had resulted from repeated cycles of deposition and erosion that had 

 passed through the cavern. One is inclined to think of caves as stable 

 repositories of sediment, slowly building up from floor to roof, but 

 typically, this is not the case with the dolomitic caves of the Sterk- 

 fontein valley. The usual situation here is that a cavern opens up- 

 ward to the surface of the hillside, while also having connections 

 downward to lower caverns. This was certainly the case at Swart- 

 krans, with the fossil-bearing sediments resting temporarily on a 

 "shelf" below a sequence of upward-leading entrances, and then 

 being carried away, in part or whole, by erosion into lower caverns. 

 The result is an intricately interwoven set of deposits of different 

 ages, each separated from the next by an erosional interval. It was, 

 in fact, when I realized that such cycles were sure to have been 



15 



