ON THE GENESIS OF THE DIAMOND 12 J 



occurrences would fall into line as phases of the same phenom- 

 enon of contact metamorphism, and to this it maybe added that 

 the, at present, striking differences between the latter and the 

 other known Brazilian occurrences would be reconcilable. 



Before leaving the topic of African analogies it may be 

 mentioned that in another Brazilian diamond region, that of the 

 river Abaiete, a porphyritic peridotite (picrite-porphyry) with 

 perofskite, quite similar to that of Kimberley as described by 

 Lewis and others, has been found. Its known occurrence is, how- 

 ever, at some distance from the diamond washings and no rela- 

 tion between the two would ever have been thought of if it had 

 not been for the Kimberley occurrence. 



In the oldest and best known of the Brazilian diamond fields, 

 that of Diamantina in Minas Geraes, there is an apparent rela- 

 tion, first noted by Eschwege in 1822 and confirmed by all sub- 

 sequent writers, between the distribution of the diamond and 

 that of the quartzose rock known as itacolumite. Eschwege who 

 first described this rock 1 recognized a schistose and a massive 

 type, the latter often presenting a conglomeratic appearance 

 and, occasionally, an apparent lack of conformability with the 

 former. 2 As, however, both types, and the schists associated 

 with the schistose one, were considered as constituting a single 

 division of the primitive group and as having a special, and non- 

 clastic, mode of origin, the two were not separated and the 

 question of the conglomeratic character was left by Eschwege 

 as an unsolved problem. The predominance of the massive 

 type of itacolumite in the Diamantina region was noted and 

 from this a genetic relation between the rock and the diamond 

 was inferred, an hypothesis which has become deeply rooted in 

 mineralogical literature. About 1840, and after the publication 

 of Eschwege's various works, diamonds were actually discovered 

 and worked in this rock at Grao Mogol, some 100 miles, more 

 or less, to the northward of Diamantina. The locality was 



1 Geognostisches Gemalde von Brasilien, und wahrscheinliches Muttergestein der 

 Diamanten, Weimar, 1822. 



2 Beitrage zur Gebirgskunde Brasiliens, Berlin, 1832, pp. 210, 216. 



