372 THE FAULT SYSTEMS IN SOUTH OF SOUTH AFRICA. 



to melt the rocks right away from the cold, solid state tu the 

 molten one, but yielding sufficient additional heat to the under- 

 lying rocks already under great heat, and pressure to make them 

 pass from the solid to the liquid condition. The gap formed 

 underground by the sinking block would act as a conduit for 

 this molten material, and make for it an easy passage to the 

 surface. This crushing, or brecciation, of the rock along the 

 fault-planes may occur in faults in which the block on one side 

 simply slides down along the fault-plane, but as far as my field 

 knowledge carries me, I do not think that it is at all common ; 

 the great faults that 1 have studied are wonck-rfully free of any 

 signs of disturbance along the fault-i)lanes. I am inclined to 

 think, therefore, that where crush conglomerates occur, it will 

 usually be found that there has been a tilting of one side of the 

 fault. Such conglomerates are found, for instance, in the 

 Table Mountain Sandstone area of Berg River Hoek, near 

 Paarl, and in the western Head at Knysna, and also constituting 

 .a sort of banket in the Tebekwe JNIine, Selukwe, in Rhodesia. 

 In the Baviaans Kloof the breccia is uncemented, each frag- 

 ment lying loose among its neighbours, but in the Berg River 

 Hoek and Tebekwe ^iine the blocks have been re-cemented by 

 deposit of silica "from circulating water, and an intensely hard 

 rock results. The originally angular blocks in the zone of 

 crushing in the Tebekwe INIine and in the Knysna Head have 

 been rounded by repeated up and down movement in the line 

 of fault, causing the loose blocks to be rolled, so that Avhat was 

 at one time a breccia has now become a conglomerate. 



Fault-planes from their nature are eminently suited for the 

 percolation of water, hence we often find springs rising in the 

 outcrops of faults, as in the case of the Cold Spring, near 

 Grahamstown. On the other hand, this ver_v fact of the break 

 in the rocks tending to fonn a water-leading carries with it the 

 consequence that prolonged percolation causes the choking- of 

 the fissure bv means of mineral deposit, and hence some fault- 

 planes are more solid and impervious to water than the rocks 

 adjoining. The c[uestion arose in the Berg River Hoek whether 

 it would be safe to build a dam wall across a particular fault 

 plane ; on investigation, it was found that where the cru.-hing 

 along the fault occurred, the rock was so firmly cemented that 

 it was far more dense than the unaffected rock in the neighbour- 

 hood. 



In the more superficial positions of the earth's crust the 

 movements that are continually occurring in it tend to tear and 

 rend it, so that underground fissures are formed. In this 

 country, where the rocks for the most part are impervious, these 

 fissures are practically the only source of deep-seated under- 

 ground water. They act practically as an artesian system in 

 that the water is conducted from a higher level to a lower, and 

 may be turned up to the surface by a bar of hard rock, sucii as 



