GEOLOGY OF THE SOUTHERN TRANSVAAL. 455 



would give them a thickness of abouh 17,000 feet, thus corro])oratinf5 

 the estimate of 18,UO0 feet obtained from the assumed rc^^ular 

 curve in the present lines of their stratification. There may be 

 nothiiii; unusual in such hollows in the granite, whether above or 

 even fur below sea-level ; but a consideration of this basin (now so 

 far inland and so high at its edges) suggests to me that the 

 Megaliesberg formation may possibly have had a marine origin, and 

 Dot necessarily have been, as hitherto generally supposed, formed at 

 the bottom of a lake at a considerable elevation. The great basin 

 may certainly have been formed by subsidence of its centre, and 

 not by upheaval around its margin ; or there may have been both 

 upheaval and depression. It is geologically and practically in- 

 teresting to determine whether the granite ridge completely circum- 

 scribes this Hand basin, and so forms a huge circular bowl filled by 

 the lower part of the Megaliesberg formation, some 60 miles across 

 and 18,000 feet deep ; or if it has an outlet or an extension carry- 

 ing these auriferous deposits into unknown areas trending towards 

 the Atlantic Ocean. 



The foregoing observations establish a minimum thickness of 

 17,000 feet for the AVitwatersrand series, uninterrupted by volcanic 

 disturbances except so far as the intrusion of some few diorite dykes 

 is concerned. To a later period of intermittent outbursts may be 

 attributed the more crystalline nature of the rocks of the upper 

 series, which encloses many interbedded trap-rocks extending over 

 large areas ; the first of these being the thick sheet of doleritc (c?, of 

 the map and sections) already referred to, which forms a distinct 

 line of demarcation. Both series of strata are persistent over very 

 large tracts, especially the auriferous conglomerates (" bankets "), — 

 a feature admitting of their identification from one side of the basin 

 to the other. This fact, with the unbroken outcrops and inward 

 dip around the basin, proves that each series of " bankets," and 

 apparently the individual deposits (with some exceptions, due to 

 thinning-out), are continuous. As regards the agency by which 

 such extensive beds of gravel, a few feet or even a few inches only 

 in thickness, can have been so evenly distributed, I confess my in- 

 ability to offer a satisfactory suggestion. 



Ferruginous shales occur amongst the sandstones (which are 

 often false-bedded and ripple-marked), most frequently between 

 what is called the " Main lleef " and the lower beds of the forma- 

 tion, one series of them invariably enclosing a band of magnetic 

 iron-ore of peculiar character and appearance. By a process of 

 chemical segregation, this iron ore has been concentrated along 

 distinct bands, black, brown, or red in colour, leaving white streaks 

 between, but not always following the lines of lamination ; the 

 banding may be straight, wavy, contorted, or folded ; bat there is 

 neither much folding of the beds nor contortion. There are two, 

 three, or more of these magnetic beds in some places, but the chief 

 one, which powerfully affects the compass, is (or certainly appears 

 to be) continuous around the end of the basin, having been traced 

 all along the Baud, round by Heidelberg, and across the Yaal Biver 



