217 



Standing just in front of the cliff at the upper part of the serra, on 

 the northern side near the Aroxi notch, is a large, high pillar, cov- 

 ered with similar rude paintings, and apparently at one time an 

 object of superstitious regard. Similar iigures are drawn on the 

 cliffs near by and in the notch. These so-called hierogh'iihics of 

 Erere were examined and copied by Wallace, but the sketches were 

 unfortunately lost. I have reproduced some of the more important 

 forms in the American Naturalisi.* Mr. J. B. Steere, on a visit 

 with me to the mountain, had the good fortune to find a large 

 fragment of silicified wood, imbedded in the sandstone, near the 

 great painted rock on top of the serra. This is clearly coniferous in 

 structure, but Dr. Dawson, to whom it has been referred, has not 

 been able to determine it. Mr. Steere also found what appears to 

 be the impression of the trunk of a large tree on the surface of a 

 bed of sandstone, on a ridge about a quarter of a mile to the south- 

 westward of the painted rock. 



One point in the geology of the Erere District is settled upon the 

 best of palaeontological evidence, and that is, the age of the beds 

 forming the great plain to the north of the serra. These are cer- 

 tainly Devonian, But what is the age of the rocks forming the 

 serra itself ? I have already expressed the opinion that the strata 

 of the serra were disturbed before the beds forming the plains 

 were laid down, since these strata are highly inclined, while the 

 Devonian rocks bordering the base are quite horizontal, presenting 

 nowhere more tlian an exceedingly slight inclination. There is no 

 reason why coniferous wood should not occur in strata of Devonian, 

 or even Upper Silurian age under the Equator; but I must freely 

 confess, that, after carefully considering the whole subject, it seems 

 to me quite probable that the Erere sandstones are really newer than 

 the fossiliferous beds of the plains, and that these last may dip 

 under the serra ; but, if this is the case, it is extraordinary that the 

 sandstones, if once continuous over the plains, should have been so 

 completely Avorn away and that the plains should have been so very 

 evenly denuded. It is also somewhat strange that the structure of 

 the serra of Tajuri should differ so markedly from that of Erere. 

 I have made a long and careful search for exposures along the base 



* Brazilian Rock Inscriptions, Amer. Nat., May, 1871. 



BUL. BUP. SOC. NAT. SCI. (28) JANUARY, 1874. 



