530 ORVILLE A. DERBY 



being equal, is determined by lapse of time, by the permea- 

 bility and by the solubility of the constituents rather than by its 

 hardness." If under permeability we include such features of 

 geological structure as fissibility, jointing, etc., which facilitate 

 the access of water to the interior of rock masses independent 

 of the structure of the rocks per se, and the position of the 

 planes of such geological structure with reference to the source 

 of water supply, the law as above stated may be taken as cover- 

 ing the entire question to be considered. 



Of the authors who have treated of rock decay in Brazil, the 

 two whose opinions are entitled to the most weight, Agassiz and 

 Hartt, believed (at least at the time of writing) in the general 

 glaciation of the country by which the element of time in the 

 formation of the present coating of decomposition products 

 would be materially reduced. If, as Branner has recently shown 

 to be probable,' these authors afterward modified their views 

 regarding glaciation, they would doubtless have expressed them- 

 selves very differently on the subject of rock decay. At all 

 events, for those who, like Branner and the writer, do not 

 believe in the glaciation of Brazil, the time element is practically 

 unlimited and the question becomes one as to whether or not 

 the other factors of permeability and solubility are sufificient to 

 account for the phenomena observed. 



That the decomposition of Brazilian rocks, particularly those 



of the crystalline and semicrystalline groups, is both widespread 



and profound is abundantly proven by the numerous examples 



cited in Branner's paper. It is even probable that the cases cited 



fall far short of the extremes that may be found when the country 



is more fully explored. Thus far, however, no authentic Brazilian 



example in gneiss equals that of the bed in the Hoosic Tunnel 



decayed to a depth of 230 feet below the outcrop (in a glaciated 



region be it noted) cited by Hunt,^ or in granite that of the 



Cornish example rotten to the depth of 600 feet, cited by 



Geikie.3 According to the latter authority, the kaolinization of 



' Jour, of Geol., Vol. I, p. 753. 



- Am. Jour. Sci., 3d ser., Vol. XXVI, p. 198. 



3 Text-Book of Geol., 2d ed., p. 322. 



