388 



ME. F. ET/TLEI ON THE OEIGTN OF [Aug. 1 894, 



In this paper the helief has already been expressed that structural 

 gradations might be traced from Arkansas stone to rocks which 

 would generally be recognized as quartzites, but it remained to be 

 shown whether such a structural series could in each instance, or in 

 every grade, be referred to a similar origin : namely, whether, at the 

 two extremes of such a series, rocks could be found wbich could be 

 proved to be siliceous replacements of limestone. 



The example which we shall next consider will, I think, help to 

 do this, since, in point of structure, it approaches more nearly to a 

 quartzite than any limestone-replacing rock with which we have yet 

 had to deal. The specimen approaching nearest to quartzite which 

 has hitherto been mentioned is the pebble from Purtiall. This 

 seems to form a link between what we recognize as tbe cryptocrys- 

 talline and microcrystalline conditions. That these conditions have 

 no distinct individuality is, I think, almost certain, since they appear 

 to depend upon the size of the crystals or grains in relation to the 

 thickness of the section, while the dark lines, which seem to be the 

 sinuous outlines of interlocking grains, when the section is examined 

 between crossed nicols, are seen to wander and frequently form ir- 

 regular contracting and expanding rings as the section is rotated. 

 That this phenomenon is due to compensation, brought about by the 

 squamose overlapping of small crystals or crystalline grains, which 

 react upon one another as diminutive quartz-wedges, seems highly 

 probable. 



Fig. 1. — Auriferous quartzite from Nondweni, Zululand. 

 (Ordinary light.) 



x 140. 

 q = quartz, c = calcite. 



Some time ago several specimens were sent me from Zululand by 

 my late sister-in-law, Mrs. Jenkinson. They were from the Zululand 

 Gold-fields, and bore no other label ; but Mr. Graham Jenkinson, 

 during a recent visit to this country, was able to give me the locality 



