THE 1--AULT S^■STI•:MS IN SOUTH OF SOUTH AFKU A. 379 



by the great trough which is found in the sea-floor ott our 

 southern coast, beginning about south of Mossel Bax- and 

 rimning due east; the other is a similar trough l)eginning some- 

 where off Port Ehzabeth and running north-east, parallel with 

 the East Coast of South Africa. Had these troughs been on 

 dry land they would have been called rift-valleys; as it is, they 

 are obscured by the difficulty of reconstructing the exact topo- 

 graphy of the sea-floor from a limited number of soundings, 

 and are only shown with somewhat hazy margins in the most 

 recent Admiralty charts. The fault-pits are always u-iost 

 extended in an east and west direction, but the east and west 

 ends, though cjuite short, are invariably parallel to the north- 

 east coast line and sea-trough. 



The flrst line of the fault-pits begins with the two basins 

 filled in with Uitenhage cretaceous deposits at Worcester and 

 Robertson. The northern margins of these two are coincident 

 with the previous fault-line which let down the Karroo beds 

 against the Malmesbury beds. The next three belonging to 

 this line also lie hard up against the foot of the Langeberg 

 Alountains; they may be designated as the Swellendam, Rivers- 

 dale, and Mossel Bay basins. The cretaceous conglomerates 

 in the Riversdale area contain an extraordinary block of melilite 

 basalt which is usually taken to be the top of a volcanic pipe, 

 but the undisturbed nature of the loosely compacted con- 

 glomerates round it, and the inclination of faint lines of strati- 

 fication in the lava, suggest that it may possibly be a giant 

 meteorite. At Mossel Bay, at Cape St. Blaize. there is a small 

 patch of cretaceous deposits to the seaward of the main basin, 

 and separated from it by an area occupied by Table Mountain 

 sandstone. This suggests that there may be a further line of 

 fault-basins off our coast concealed under the Avater. The 

 main line of the fault-pits belonging to what I have called the 

 first series is continued by the basins at Knysna and Pletten- 

 berg's Bay, the latter being double, one on the coast running 

 down the Pisang River, and the conglomerates in it jutting out 

 to sea in the fine headland of Seal Poiiit. and the other further 

 inland, l)'in-g along the Bitou River. 



The second or middle line of fault j^its begins with the 

 great basin of Oudtshoorn, along which runs the Olifants River; 

 it is constricted in the nfiddle by a northerly spur of the Kam- 

 manassie Mountains. Plere also the northern liordor of the 

 fault-pit is coincident with the ])revious fault-line, which lets 

 down the Bokkeveld beds against the Cango beds. The line 

 is continued by five smaller basins along the Baviaans Kiver, 

 and ends in the l)asin at the mouth of the ( iamtoos Iviver. 



The third or inmost line is represented b\' the wide area of 

 Uitenhage beds around Uitenhage ; it is made u]) of at least 

 three smaller parallel troughs, which are con.cealed by the 

 cretaceous beds which have covered the smaller details. Tliis 



