REPORT OF THE DIRECTOR I9I4 I25 



The composition of this austral Devonic fauna> on close analysis, 

 brings out the evident fact that, whatever its origin, it has developed 

 its peculiar characteristics under the influence of isolation from the 

 other Devonic basins and shelf-seas of this stage. In South Africa 

 it may be demonstrable on further evidence that the same fauna 

 was preceded by a period of continental deposition properly included 

 within the Devonic system. But in either case there is no indication 

 that the primary calcareous term of the northern Devonic, by its 

 absence in the south, is to be regarded as a factor of any weight 

 whatsoever in estimating the relative stage of the austral fauna. In 

 the Falkland islands, where the affiliations of the marine fauna are 

 more intimate with that of the Bokkeveld beds than with the nearer 

 South America fauna, there fails as yet any evidence of a pre- 

 liminary Devonic deposit of continental character. 



These extensive and reasonably profuse faunas of the southern 

 Devonic strands developed along a continent obviously separated 

 from that at the north, and for most of its extent widely by equa- 

 torial Atlantic waters, but narrowly in the subequatorial latitudes 

 of Brazil, where the Amazonas faunas, with affiliations toward the 

 north, lie not far away from the Devonic beds of Matto Grosso with 

 evident alliances with the south. The marine Devonic was the 

 strand of a Pre-Gondwana land of whose constituent sedimentary 

 rocks we know little save for the occasional dredging of altered 

 sediments from the Atlantic bottom, the gneissoid inclusions in the 

 deep-seated lavas of the mid-Atlantic islands, and perhaps some 

 part of the crystallines of the southern islands. South Georgia, the 

 South Shetlands and of Antarctica. To these are to be added the 

 South African premarine Devonic series capped by the Table 

 Mountain sandstone, now regarded by some writers as of glacial 

 origin, and, of course, some part of the Precambric crystallines of 

 South America and Africa. 



Our present knowledge leads us to the conception of an Andean 

 Siluric and Cambric land reaching far to the north along the 

 Cordilleran rib, for the limestones of easterri Argentina, Bolivia and 

 Peru carrying Liorhynchus bodenbenderi and its associated fossils, 

 are Siluric, not Devonic. The stretch of the Devonic strand along 

 the Cordilleran rib far to the north is now well known and there is 

 every reason to believe that the Pre-Gondwana land which traversed 

 the south Atlantic extended an arm well to the north on the Pacific 

 side. Pre-Gondwana land was thus a very ancient austral con- 

 tinent, and wholly comparable in extent and age to Laurentia at the 

 north. The latter, in days before the Caledonian folding, traversed 



