ON THE GENESIS OF THE DIAMOND 13$ 



tified the original rock type as hornblende-schist, 1 giving the 

 impression that the diamond occurred throughout the whole 

 mass. The limitation of the diamond to certain streaks, or lay- 

 ers, was recognized by Burton, who gives the following excel- 

 lent description of them. 



"The richest lode (corfio) is No. 3, or the highest. The 

 strike of the ribboned clays is north and south, bending [dip- 

 ping] eastward. The lode inclines towards the higher grounds, 

 and thus the owner hopes to find the gem-bearing strata spread- 

 ing over the crest or watershed ridge which forms his property.. 

 Through the ferruginous sandstone (borra) and the white feld- 

 spathic matter run dikes and lines of fragmentary rock crystal, 

 sometimes fibrous like arragonite, and often finely comminuted. 

 Large pieces of imperfect specular iron and thin strata of quartz,, 

 yellow and brown at the junction, thread the argile, and I was 

 shown a specimen of fine sandy conglomerate, blackened and 

 scorified by the injection of melted matter. 2 The characteristics 

 of this upper lode are a dryer clay, silica, a trace of copper [a 

 green silicate of the nature of chlorite or serpentine?], of iron 

 cement, and of Canga in small pieces ; when the specular iron is 

 in large pieces and abundant the rock is rich in gems. Its 

 'agulhas' [rutile] are iron-like bundles of needles welded 

 together by intense heat; some are double, the fibers coming at 

 obtuse angles. The 'Agulhas Cor de Ouro' have a burnished 

 coppery surface, whence the name. Throughout all these corpos 

 the diamonds are small, averaging perhaps a little under one 

 grain or 64-72 per oitava ; they are mostly crusted superficially 



1 Although Rose failed to find hornblende in the material submitted to him by 

 Heusser and Claraz, it is possible that this idea was not entirely without foundation.. 

 Throughout the whole region traversed by them, intercalations and dikes of amphi- 

 bolic and pyroxenic rocks are frequent in the schist series and generally they are the 

 only ones of the not distinctly quartzose rocks whose original composition can be 

 readily determined. As they usually give green decomposition products the staining 

 referred by Burton to copper may, with considerable plausibility, be attributed to^ 

 them. To judge from other exposures, the absence rather than the presence of such 

 rocks at Sao Joao da Chapada would be a motive for remark. 



2 Probably a " Canga " or mass of sand or clay cemented by limonite common in 

 such deposits. 



