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



opinion expressed b}'' Julien,^ that the simple heating and cooling of 

 the rock is not, at the temperature it undergoes, a matter of much im- 

 portance. The importance of such heating and cooling comes almost 

 entirely from the moisture of the atmosphere in which it takes place and 

 from the train of disasters which follow. 



What is true of the penetration of heat into the rock mass on a plane 

 surface as compared with that penetrating it on an angle is true also of 

 the penetration of other agents of disintegration, such as carbonic acid 

 and water.f This is shown by the fact that the process of concentric 

 decay goes on beneath the surface as well or even better than above 

 ground. Julien found evidences of the penetration of surface waters in 

 micaceous gneiss to a depth of about 10 feet. J 



The most important contribution of changes of temperature, therefore^ 

 toward rock decomposition is in the opening of crevices for the admis- 

 sion of water and acids which attack the constituent minerals. 



Again, exfoliation, whether of mountain masses or of boulders, can 

 take place only with massive and homogeneous rocks. Jointed, bedded, 

 or heterogeneous bodies would admit agents of decomposition unevenly 

 and permit the formation of grooves. 



Some reference should be made here to the theory proposed 1)}^ von 

 Eschwege, who says : § 



"The formation of round masses by concentric exfoliation, . . . especially 

 in granite, granite-gneiss and greenstone diabase, depends upon attraction (central- 

 kriiften) and the effects of affinity, just as does the formation of strata and as does- 

 every regular crystallization. While the entire mass is influenced by the principal 

 force of cohesion which affects the inner layers (centralfachen), this force is also 

 distributed to a small degree along single i)oints in the layers, and exercises its 

 power of attraction in all directions as far outward as its force is not nullified nor 

 hemmed in by the points lying next to it. The nearer it is to the center of this 

 force, so much firmer a nucleus does the rock form ; the further away from it it is^ 

 so much looser and more scaly does the mass surrounding the nucleus become, and 

 finally, becoming still looser, it passes into a neutral condition until the expression 

 of the force of the next point is seen in the formation of a new globe. In this 

 manner, according as these centers are close together or far apart or have a strong 

 or weak effect, the spheres are larger or smaller, and when of granite (in Portugal, 

 for example, in the neighborhood of Oporto and Penafiel, in the province of Minho), 

 are at times 10, 20 or even 50 feet in diameter; they exist also in oolite of sec- 

 ondary formation as very small spheres. 



" Our great Werner also considers spheroidal forms, especially those of basalt, 

 as orignal and not as having come from weathering, as several have concluded. 



* A study of the New York obelisk as a decayed boulder. A. A. Julien. Ann. N. Y. Acad. Sci., 

 vol. viii, 1893, pp. 93-166. 



t For the mathematics of exfoliation see Monograph xiii, U. S. Geo). Survey. Washington, 1888, 

 Geolog3' of the Quicksilver Deposits of the Pacific Slope. G. F. Becker. Pp. 68-72. 



X Proc. Am. Assoc. Adv. Sci., vol. xxviii, 1879, p. 377. 



2 Beitnlge zur Gebirgskunde Brasilien. W. L. von Eschwege. Wien, 1832, pp. 14, 15. 



