492 Annals of the South African Museum. 



sufficiently important to warrant their promotion to a system of 

 their own, they should he grouped with the Uitenhage System. 

 His reasons for this view, as stated in Proc. Geol. Soc. S. Africa 

 1919, are open to question. "The most important of all fauna! 

 changes, the passage of the vertebrate remains from reptilian to 

 mammalian" is not yet proved to he a feature of the Stormberg 

 Beds; the fauna is a reptilian one, with members of the Beaufort 

 Theriodontia still living; the Dinosaurs are, as we have seen, Triassic 

 and not Jurassic nor Cretaceous forms; and Glossopteris is still a 

 feature of the Molteno Beds. 



Important evidence bearing on the age of the sediments has 

 recently been forthcoming from South America. The beds of Karroo 

 age in the State of Parana, Brazil, show the following succession :- 



6. Eruptivas da Serra Geral 600 metres. 



5. Arenito de Botucati 200 ,, 



4. Serie Rio Rasto 100 



I Calcareo Rocinha .... 3 ,, 



3. Serie Passa Dois Grupo Estrada Nova . . . 150 ,, 



r Grupo Iraty 70 ,, 



2. Serie Tubarao 180 



1. Serie Itarare 350 ,, 



The Itarare Series is glacial, and its upper portion contains a marine 

 fauna of Lamellibranchs and Gasteropods. 



The Tubarao Series contains a Glossopteris flora. 



The Iraty Group contains the reptiles Mesosaurus and Stereosternum. 



The Rio Rasto group has yielded the reptiles S'caphoni/x fischeri and 

 Erythrosuchus sp. and is presumably approximately equivalent to our 

 Upper Beaufort Beds. Recently Holdhaus (1918) has described from 

 the same Series the lamellibranchs Solenomorpha similis, S. intermedia, 

 S. altissima, S. deflexa, and Sanguinolites elongates - - an assemblage 

 which has led him to class the beds definitely as Permian. 



The Botucati sandstone, sandwiched as it is between the beds con- 

 taining Ert/throsuchus and the thick volcanic outpourings which, like 

 those of the Stormberg, made an end of sedimentation, can but cor- 

 respond to the Stormberg sediments of South Africa. Following as 

 it does conformably upon strata now classed as Permian there can 

 be little doubt that it is not later than Triassic in age a result 

 in accordance with that now obtained from our consideration of the 

 South African deposits. 



