CLIMATIC RELATIONS 197 



the region of the Catinga limestones, on the other hand, we have stream 

 channels now in the process of being filled up with deposits of lime rock ; 

 over their bottoms, and we have stream channels and narrow valleys being 

 encroached on by the inbnilding of travertine banks from one or from 

 both sides. 



It is evident that there was a time not long ago when these same rivers 

 did cut their channels; it is evident, too, that this cutting process, for 

 some reason, has ceased, or has become so enfeebled that the actual cutting 

 may be neglected. 



The only way I have been able to account for this change in the char- 

 acter of the work being done by the streams is on the theory of a change 

 in the amount of the rainfall of the region. Such a change could be 

 brought about by a difference in the altitude of the land with reference 

 to the sealevel. 



The east coast of Brazil stood considerably higher during the Miocene 

 than it does at the present.^ The evidence of this Miocene elevation can 

 not be repeated here, but it may be said that it is conclusive in regard to 

 the time and character of the movement that the region affected included 

 the whole coast of Bahia, Sergipe, Alagoas, Pernambuco, and probably all 

 of northeastern Brazil, but that, so far as presented, the evidence is de- 

 fective in regard to the amount of the uplift. It is clear, however, that a 

 slight elevation, or one of a few meters only, or even of a few hundred 

 meters, could have no considerable effect on the rainfall. The size and 

 character of the abandoned and choked up river channels in the limestone 

 region of the interior of Bahia suggest that the Chapada Diamentina was 

 high enough to produce a considerably larger rainfall than we now have — 

 quite enough to enable the streams not only to keep their channels open, 

 but to deepen them as we should expect in an elevated region of heavy 

 intermittent tropical rains. This period of elevation was a period of 

 much greater precipitation and of much greater activity on the part of 

 the streams, both in the higher and in the lower lands. 



The greater rainfall of the period of elevation must have favored a 

 ranker vegetation and a more rapid decay of vegetation, a consequent 

 increase of carbon dioxide and an increased chemical, as well as me- 

 chanical, removal of the limestone over the entire area. 



The geography of the region and especially the relations of the locality 

 to the prevailing winds and to the ocean are factors of prime importance 

 in this case. The winds set uniformly from the ocean toward the interior 

 along this part of the continent,^^ and blowing from the warm ocean into 



» .T. C. Branner : The stone reefs of Brazil, etc. Bull. Mus. Comp. Zool., vol. xliv, 

 Cambridge, 1904, pp. 127-147. 



10 E. Mouchez : Les cotes du Bresil, 2d ed. Paris, 1876 ; plates following p. 272. 



