58 ' THE KHEIS SERIES 



certain that they are partly of sedimentary and partly 

 of volcanic origin, and that they have been invaded by 

 granite and sheared during, and subsequently to, the 

 intrusion. The variety of schists and granulites pro- 

 duced by the changes due to the influence of the granite 

 and pressure indicates original differences in composi- 

 tion ; for example, the abundance of garnet and epidote 

 in a rock that was evidently a lava shows that the rock 

 had lost much of its alkalis before it was altered. It is 

 very probable that the transference of felspathic sub- 

 stance from the granite to the rocks invaded is the 

 cause of the difficulty in the delimitation of the quart- 

 zites and granulitic gneiss. 



The whole thickness of the Marydale beds may not 

 be much more than 1,500 feet, though in some localities 

 repetition of similar beds by faulting and folding allows 

 the formation to occupy a belt five miles wide. 



The resemblance of the Marydale beds to the Kraaipan 

 formation is extremely close, though the proportion of 

 banded ironstones and cherts to volcanic rocks is greater 

 in the latter. The Marydale beds have been invaded by 

 granite, but the Kraaipan formation appears to be 

 younger than the Bechuanaland granite. It is im- 

 portant to note, however, that there are quartz-felspar 

 grits in the Marydale beds and that the Kaaien beds 

 were evidently derived from granitic rocks, so although 

 the Kheis series were invaded by the granite of Prieska 

 and Upington they,' like the Kraaipan formation, may 

 be younger than the Bechuanaland granite. 



In the Marydale beds there are some intrusions of 

 serpentine older than the granite, for example at Zwart- 



