FAUNAL PROVINCES AND REALMS 345 



carinalus, Leptocoelia flabelliteSj Spirifer of the antarcticus group, and 

 primitive but large terebratulids like Rensselceria, Scaphioccelia, etcetera. 



All of these local faunas have a common generic and specific develop- 

 ment, and because this exj)ression is quite unlike that of the same age in 

 North America and Europe, Clarke has distinguished them as of the 

 austral Lower Devonian realm, '^as an emphatic distinction from the 

 boreal faunas" of the Northern Hemisphere. 



When we examine into the geographic relationships of these various 

 local faunas of the austral realm, it is seen that those of Bolivia, Peru, 

 and Argentina are very similar and more closely related to those of 

 Parana and Matto Grosso. Those of northern Brazil, however, are quite 

 different, and, even though there is here a mixed boreal and austral de- 

 . velopment, on the whole the Amazonian Devonian is clearly of the austral 

 realm. On the other hand, the fauna of the Falklands is already more 

 closely related to those of South Africa, even though they are more than 

 4,000 miles aAvay, while it is only 1,500 to 2,000 miles to the nearest 

 South American assemblage. 



The continent Gondwana. — Now let us look into the paleogeographic 

 significance of these austral Lower Devonian faunas. The geography of 

 the places recited shows them to lie on or toward the margins of an ex- 

 tensive transverse continent formerly extending from western South 

 America across Brazil and the medial or tropical Atlantic Ocean to east- 

 ern Africa, a land that has long been known as Gondwana. If there had 

 been in Devonian times a tropica] Atlantic like that of today, the African 

 faunas would be very dissimilar from those of South America, and there 

 would have been developed two austral faunal realms instead of one, and 

 this through isolation and shallow-water migrations around two conti- 

 nents instead of one. 



Nevertheless, to make sure of our conclusions that Gondwana across 

 the Atlantic is a fact, and the cause for the existence of but one austral 

 Ijower Devonian realm, avc get further and complete verification in the 

 distribution of the marine Pennsylvanian, and especially the Jurassic and 

 Comanchean, faunas of western South America. All of these assemblages 

 constantly have forms in common with the Mediterranean countries, and 

 chiefly with their souther]i or African extensions, while but very few 

 forms indeed are in common with the Indian Ocean. Hence we see that 

 during these times there was no equatorial Atlantic, l)ut, on the contrary, 

 that there was a land liere along wliose northern and western strands the 

 sliallow-water life migrated Ijack and fortli. All of this was plain to the 

 great Neumayr, who as early as 1885 traced out tlie distribution of the 



