434 Lisboa — Permian Geology of Northern Brazil. 



the "grota do Mendes' 1 which runs from that plateau to the 

 Parnahyba. This locality is in the channel of the Parnah} 7 ba 

 river between Floriano and Nova York. 



Shales. — The sandy Permian formation in the section exam- 

 ined rests upon ashen gray marly shales and calcareous beds, 

 and is capped by a thin layer of conglomerate. A thin yellow 

 residual intercalated rock seems to be derived from a limestone. 

 These rocks outcrop in the " grota do Mendes " and form the 

 bed of the Parnahyba river in that region. They are there 

 covered by a gray sandstone, but they were not seen elsewhere. 



The Mendes sandstone. — The Mendes sandstone overlies the 

 shales. It is an ashy gray sandstone grading to white in its 

 upper part. It outcrops at the level of the banks of the Parna- 

 hyba, and contains in the " grota do Mendes " a great abun- 

 dance of Psaronitis and conifers which were collected from 

 the same bed. Some of them have the same dark color as the 

 rocks. The upper part of this sandstone is white, friable, and 

 is probably the rock which forms the marginal plains of the 

 Parnahyba, down stream to Floriano and beyond. 



Along the roads overlying this rock and forming the surface 

 is a recent conglomerate, containing trunks of dicotyledonous 

 plants derived from a later sandstone of the plateaus. 



The red sandstone of Jaboti. — The Jaboti sandstone is red 

 with purple spots. It overlies the formation just described. 

 The large and beautiful specimens of Psaronius which I col- 

 lected are from the low plateau of Jaboti. Some of them are 

 forty centimeters in diameter, and beautifully preserved. I 

 collected in situ part of a trunk, which I preserved, incrusted 

 by the sandstone of the plateau. The other three specimens 

 came from this same trunk, but the material in which they were 

 imbedded had been removed before my visit. 



On examining the material from a well which was being dug 

 at this time in Floriano, I noticed that a layer of sandy shales of 

 a chocolate-red color separated the gray sandstone of Mendes 

 with six meters of thickness from the red sandstone of Jaboti. 

 Although these appeared to be distinct, it seems best to include 

 them with the Jaboti sandstone because it appears that these 

 beds of shale occur frequently in other localities intercalated 

 in this red Jaboti sandstone, which has a great development, 

 and extends north and south forming the second bottoms of 

 the Parnahyba river. 



In Amarante the river cuts a range of plateaus with an 

 escarpment 150 meters high. They are made up of white and 

 red sandstone with pisolitic concretions. In the lower part 

 they contain flows of normal diabase. 



The chapadas which are seen farther on in Castelhana are of 

 the Jaboti sandstone. This same sandy formation continues 



