332 SCIENTIFIC RECORD FOR 1882. 



conformably upon the older crystalline rocks, and consisting in great part 

 of quartzites, often granular and sometimes flexible, constituting the 

 so-called itacolumite, with unctuous talcoid schists, containing hydrous 

 micas, chloritic and argillite beds, specular schistose iron-ore (itabirite), 

 and great masses of crystalline limestone. The analogies between this 

 Brazilian series and th*' Taconian or Itacolumite series, as studied by Lie- 

 ber in South Carolina, were long since pointed out by him, and the close 

 resemblance between a collection of these rocks from the province of 

 Miuas-Geraes, in Brazil, where they are largely develoj)ed, and the Ta- 

 conian in Pennsylvania, was later insisted upon by Hunt. The ancient 

 series in Brazil has afforded no organic remains, but being unconform- 

 ably overlaid by older Paleozoic rocks is conjectured by Derby to be 

 altered Cambrian. 



This Itacolumite series is the source of the diamonds of Brazil, as 

 shown by Gorceix and by Derby. While these gems are also met with in 

 secondary and derived rocks, they are found in the district of Diaman- 

 tina, in certain unctuous banded clays which are seen to be derived by 

 subaerial decay from schistose beds belonging to the Itacolumite series, 

 •which here has a considerable dip to the eastward. These clays are of 

 various colors (red, white, or black), from disseminated iron-oxide. 

 Associated with the diamonds in the clays are tourmalines, rutile, ana- 

 tase, martite, and oligist, most of which minerals have been found in 

 quartz veins, with pyrites and gold, traversing the Itacolumite series. 

 Similar clays, derived from the decay of the accompanying schists, are 

 found with the Taconic quartzites and limestones throughout the Apal- 

 lachian valley. It is in these that chiefly occur the deposits of limon- 

 ite derived by epigenesis both from siderite and from j)yrite, which are 

 so extensively mined from Vermont to Alabama. 



A similar Itacolumite series is seen in Guiana, and has been compared 

 by J-annetaz with that of Brazil. Eocks apparently the same occur 

 in the island of Trinidad, where they were many years since studied 

 by Wall and Sawkins, who described them as the Caribbean group. 

 Since this they have been examined by Guppy, and more recently by 

 Crosby. This series, which is, according to him, not less than 10,000 

 feet thick, consists of quartzites, with argillitesand hydrous mica schists, 

 with a great body of crystalline limestone or marble, sometimes mica- 

 ceous, succeeded by argillites, hydrous mica schists, and sandstones, the 

 whole, according to Crosby, strongly resembling the Taconian as seen 

 in ^Massachusetts. Overlying nnconformably this ancient series, which 

 appears to be unfossiliferous, is a dark-colored, com])act, fossiliferous 

 limestone, with interbedded shales, in which, among many obscure forms, 

 Gui)py recognized Murcldsonia Anna and M. linearis, both found in the 

 Calciferous sand-rock in Canada. 



Analogies both stratigraphical and lithologicnl serve to connect the 

 Caribbean group of Trindad with the Taconian and Itacolumite group 

 of North and South Amewca, and to assign to all these a position below 



