260 J. C. BRANNER — DECOMPOSITION OF ROCKS IN BRAZIL. 



Minas that there are but few rocks uncovered by soil. The famous terra 

 roxa of Sao Paulo is derived by decomposition from the igneous rocks that 

 cover a large part of that state. 



A. Pissis says that in the Serra de Goitacazes the gneiss is changed to 

 a reddish clay to a depth often of more than 100 meters.* 



In the highlands of Brazil decomposition is so widespread that it is 

 often impossible to find enough exposure of hard rock in place to work 

 out the structure. Through the granite and gneiss regions this decom- 

 position may be seen in the railwa}^ cuts from Rio de Janeiro across the 

 Serra do Mar and the Serra da Mantiqueira, and this same decomposition 

 is common wherever the granites and gneisses occur through the interior^ 

 though it is not to be inferred that there are no exposures of hard rock. 



The other rocks of the highlands are mostly metamorphosed, namelyy 

 schists, itacolumites, iiahintes, jacu tin gas and a recent surface formation 

 of iron cement known as canga. Occasionally these rocks, especially the 

 itacolumites, stand out as bare and rugged mountains,t but over a large 

 part of region the geology is masked by a thick coating of soil, and de- 

 composition has profoundly affected the rocks in place. Gerber notes 

 that the gneiss is specially subject to decomposition. J 



James E. Mills, who lived and traveled in Brazil more than a year, in 

 speaking of the province of Minas Geraes says : 



" The gneiss and slates are softened to great depths from the surface. I have 

 seen sections showing a thickness of over 100 feet (estimated with the eye) of this 

 softened rock, and yet not reaching to the bottom of it." § 



To the traveler in Minas Geraes one of the striking sights is the enor- 

 mous gullies through which the roads and mule trails often pass. These 

 gullies are always on hillsides, though not necessarily high or steep ones, 

 and have frequently been made by the washing out of the soft mud from 

 the bottoms of the paths. Once the natural surface is broken by a path, 

 the heavy rains rapidly deepen the channel, and the tropelros continue to 

 follow the old road, which sinks year after year into the earth. These 

 gullies are alwa3^s V-shaped and are often so narrow in the bottom that 

 two loaded mules cannot pass each other in the path. 



* La position g6ologique des terrains de la partie australe du Br6sil. M. A. Pissis. M<^moire de 

 rinst. de France, x, 1842, p. 358. Hartt erroneously makes the depth of the decomposition given by 

 Pissis 300 meters instead of 100. Geology and Phys. Geog. of Brazil. Oh. Fred. Hartt. p. 25. 



fHussak in his Relatorio Parcial, 114, notes the greater resistance of itacolumite than of sohists* 

 while Heusser and Claraz express the opinion that " metamorphic schists and itacolumite are very 

 liable to decomposition. The decomposition of itacolumite, which is essentially quartzose, is es- 

 pecially characterized by a Assuring of the rock, which falls to powder." Gi.sement et exploitation 

 du diamant dans la Province Minas Geraes au Bresil. Ch. Heusser et G. Claraz. Ann. des Mines,^ 

 6me s6r., xvii, p. 291, 



X No^oes da provincia de Minas Geraes. Henrique Gerber, 2d ed. Hanover, 1874, 18. Derby 

 speaks of the difficultj'^ of working out structure in the interior where decomposition is so deep, 

 Archivos do Museu Nacional, iv, 1881, p. 125. 



§ Quaternary deposits, etcetera. James E. Mills. American Geologist, vol. iii, June, 1889, p. 351, 



