INEQUALITY OF THE ROCK DECOMPOSITION. 265 



the drift (much of this decayed material Agassiz regarded as of glacial 

 origin) as attaining a thickness of 162 feet.* 



Absence of decomposition. — It is worthy of note that in certain arid re- 

 gions of Brazil decomposition has not been nearly so deep as it has been 

 in the forest-covered parts. This is a striking feature of surface geology 

 in the Cretaceous and Tertiary regions of northeastern Brazil. 



Beginning in the high Tertiar}^ plateau of the interior of Bahia and in- 

 cluding much of Sergipe, Alagoas,t Pernambuco, Parahyba, Rio Grande 

 do Norte and Ceara, X the soil is, for Brazil, remarkably thin in many 

 places, especially in the more elevated campos, where the rocks are clayey 

 and the drainage is rapid. Spix and Martins must have been impressed 

 by this fact, for they were of the opinion that the soil had been removed 

 by wave-action {Meerfluthen) from much of the area. § 



The campo region of the Tertiary rocks about Erere and Monte Alegre, 

 on the Amazon, Hartt found destitute of soil. || 



One of the noteworthy features of this decomposition, wherever it 

 occurs in Brazil, is that it does not penetrate the rocks everywhere alike, 

 even when they are massive and apparently homogeneous. 



Mr Darwin noted that the " decomposition did not appear at all con- 

 formable with the present undulations of the surface," ^ 



Decomposition proceeds along joints and other planes of weakness, 

 and as it penetrates to greater depths undecayed masses are left behind 

 in the shape of boulders of decomposition. 



I have no doubt that decay is greatly hastened by the localization of 

 conditions favoring rock decay, and in the main the generalizations of 

 Pumpelly** and of Gilbert ft ^^ regard to the action of plants hold 

 good, though there are many exceptions to such a rule. 



The unequal resistance of certain great bands of gneiss is well illus- 

 trated in the flat-sided peaks about Theresopolis, which form the organ 

 pipes of the Organ mountains. 



In deep mines and tunnels this selective action is shown by occasional 

 beds of soft materials in the midst of hard ones. In the Morro Velho 

 mines such a soft bed was struck at a depth of 755 feet, after the shaft 



*Sur la geologie de I'Amazon. MM Agassiz et Coutinho, Bui. de la Soc. Geol. de France, 

 1867-"68, XXV, p. 687. 



fSee also Der Sertao der Provinz Alagoas u. Die Falle des Paulo Affonso. Rio de Janeiro, 1880, 

 pp. 30, 31. 



I Trabalhos da Commissao Scientifiea, i. Rio de Janeiro, 1862. Rel. da Sec^ao Geologica. G. S. 

 de Capanema, cxxv. 



§ Raise in Brasilien, iii, 1873. 



I! Contributions to the geology and physical geography of the Lower Amazonas. Ch. Fred. 

 Hartt. Bui. Buffalo Soc. Nat. Sci., 1874, p. 211. 



II Geological Observations, p. 428. 



** Amer. Jour. Sci., 3d ser., vol. xvii, 1879, p. 137. 

 +t Geology of the Henry Mountains, p. 119. 



