246 ANNUAL OF SCIENTIFIC DISCOVERY. 



been found by Mr. Huxwell and others at different localities, 

 among other places at a point on the Maraiioji, 30 miles below 

 Pebas. They seem to be chiefly of species inhabiting brackish 

 waters. 



Prof. Hartt also differs from Prof. Agassiz in his interpretation 

 of the Amazonian geology. While recognizing with him the 

 former existence of glaciers under the tropics, he considers the 

 clay and sandstone deposits as tertiary in age, and as contempo- 

 raneous with similar formations in other parts of Brazil where 

 they are overlaid by the true drift. 



BRAZILIAN GEOLOGY. 



The eozoic rocks are represented by gneisses, of which the dis- 

 tribution is very general and the extent enormous. As a rule the 

 gneiss, which on the east is coarse and porphyritic, becomes 

 finer as one proceeds westward, and is generally overlaid by mica 

 slates or mica-schistose gneiss. This is certainly the oldest rock 

 formation of the Brazilian plateau. Dr. Hunt has examined a 

 large series of rocks brought from Brazil by Prof. Hartt, and 

 finds that they are very similar in character to the Laurentian 

 rocks of North America. With this formation the Brazilian 

 gneisses also agree in the presence of interstratified beds of 

 limestone and in the absence of clay-slates. The rocks in the 

 northern provinces have been but slightly examined, and the study 

 of those of the southern provinces is attended with difficulty on 

 account of the extent to which the rocks are decomposed. The 

 exact succession of the different members of the metamorpliic 

 series lying just inside of the gneiss belt has not been satisfacto- 

 rily worked out. The clay and talcose schists, the itacolumite, 

 itaribite, and other associated metamorphic rocks of this region 

 appear to be lower paleeozoic in age. The gold-bearing rocks 'in 

 Minas Geraes resemble the similar auriferous series of the south- 

 ern Atlantic States, in which itacolumite occurs. The meta- 

 morphism has been so extensive that all traces . of fossils have 

 been obliterated. Most of these rocks are probably Silurian, al- 

 though some of them may be Devonian. 



The carboniferous rocks are developed to a slight extent in the 

 southern part of Brazil, the coal-beds being a coast formation, but 

 slightly disturbed and of a bituminous character. The triassio is 

 represented by a series of red sandstones, lithologically identical 

 with the Connecticut River sandstones. There is no trap asso- 

 ciated with them as far as yet observed. The cretaceous rocks 

 begin a few miles south of the Bay of Bahia, and extend at inter- 

 vals along the coast to the northward. It is difficult to estimate 

 the exact extent, as they are covered by tertiary beds. They 

 probably underlie the tertiary deposits of the whole valley of the 

 Amazons. The clays and ferruginous sandstones forming the coast 

 plains overlying unconformably the cretaceous and covered by 

 the drift clays, Professor Hartt refers to the tertiary, although no 

 fossils have been found. To the same formation, though of later 



