Lisboa — Permian Geology of Northern Brazil. 431 



to the fossil. This erroneous identification of the woody cen- 

 tral roots of Psaronius with those of the fibro-vascular bun- 

 dles of the carnaubeira {Copernicea cerifera), abundant in the 

 region, not only shows the observant character of the inhabit- 

 ants, but also affords a relatively safe basis for the provisional 

 identification of fossil woods occurring in localities not 

 examined. 



5. Fossil fishes. — In the Pastos-Bons region, according to 

 information furnished by the inhabitants, there are fossil 

 fishes. I was given, in that locality, a piece of the backbone 

 of a fish, as yet unclassified, but I have no information in 

 regard to the character of the beds from which it comes. 



In this zone there should occur the contacts of two or three 

 series of sediments, as will be seen farther on. The identifica- 

 tion of this fossil will probably determine the geological age 

 of the sediments of Pastos-Bons, and further study will permit 

 a better discrimination of the beds. 



6. Pish scales. — In Floriano I got some ganoid fish scales 

 that came from Frasqueira, seventy-two kilometers from Flori- 

 ano, but the locality could not be examined. 



Paleobotany. — All the specimens of fossil plants found were 

 silicified, and at times they were beautifully preserved. Dr. 

 Orville A. Derby submitted the specimens we collected to 

 Count Salms-Laubach, who made a detailed paleobotanical 

 study of them and pronounced them to be Permian types. 

 This study gave evidence of the existence of various types 

 of Psaronius. It is to be noted that specimens of Psaronius 

 are still found in place in the purple sandstone of the Jaboti 

 tablelands above the gray sandstone containing conifers and 

 Psaronius of other types. 



These were in the base of sandy sediments which overlie 

 marly shales containing calcareous layers. 



From the erosion of the gray sandstone beds there resulted 

 the accumulation of the silicified plants on the surface of the 

 lower slaty shales. The discovery of a great number of these 

 localities at considerable distances apart, made in a hurried 

 reconnaissance and the presumption that different types cor- 

 respond to different localities, as well as the incomplete study 

 of poorly preserved specimens, emphasize the importance, 

 from a paleobotanical point of view, of the study of the Per- 

 mian plants of northern Brazil. The great extent of these 

 sediments in the north is also noteworthy. 



In every locality where conifers were found in place I have 

 found that they either accompany the Psaronius and come 

 from the same sedimentary deposit, or else were isolated and 

 not accompanied by other fossil plants. 



In no undoubted Cretaceous locality did I find conifers asso- 



Am. Jour. Sci. -Fourth Series, Vol. XXXVII, No. 221.— -May, 1914. 

 30 



