PRESENT PROBLEMS 579 



sphere, namely, Africa, South America, and Australia, were inter- 

 mittently, but not continuously connected by land. A great southerly 

 continent, Notogsea (Huxley, 1868), connected with a south polar 

 Antarctica, now submerged, is a theory very widely supported by 

 zoologists 1 and, I believe, by botanists, although its existence is still 

 denied by certain geographers (Murray). We find Permian, Jurassic, 

 late Cretaceous, and early Tertiary proofs of Antarctica in the fresh- 

 water crustaceans (Ortmann), in fresh-water fishes (Gill), in littoral 

 mollusca (Ortmann), in reptiles (Smith Woodward and Osborn), in 

 birds (Forbes and Milne Edwards), in worms (Beddard), in the Aus- 

 tralian animals (Spencer), in the fossil mollusca of Patagonia (Ort- 

 mann), and in the fossil mammals of Patagonia (Ameghino). To 

 marshal and critically examine all this evidence and convert this 

 most convenient Antarctic hypothesis into an established working 

 theory I consider one of the most pressing problems of the day. 



Problem of the Source of the Reptiles and Mammals 



Returning from this geographical detour to paleontology as history, 

 we should first note that already in the Permian there was developed 

 such an astonishing variety and differentiation of the reptiles that we 

 must look to future discoveries in the Carboniferous to find the actual 

 points of descent of reptiles from the amphibia. These Permian and 

 Lower Triassic reptiles 2 are of three kinds, comparable to a parent 

 (Cotylosauria) and two offspring (Anomodontia and Diaptosauria) . 

 In the parent group (the Cotylosauria, or solid-skulled reptiles), we 

 find so many fundamental similarities to the Stegocephalia, or solid- 

 skulled amphibia, that only by the possession of many parts of the 

 body can we surely distinguish reptile from amphibian remains. The 

 primordial reptile was probably altogether a land animal continuously 

 using its limbs in awkward progression, bringing forth its young by 

 land-laid eggs and probably possessing gills only as vestiges. These 

 cotylosaurs show very wide geographical distribution, South Africa, 

 Siberia, Great Britain, and North America, and equally remarkable 

 adaptive radiations of habit into small and large, horned and hornless 

 types, some of which were certainly dying-out branches, while others 

 led into the two offspring groups. 



Leaving this parental order, in the Permian and Lower Trias, we 

 first see in the older offspring, the Anomodontia, reptiles of varied 

 size and description, carnivorous and herbivorous in habit, most 

 abundantly found in South Africa, in Asia, and in Europe, and not 



1 A. E. Ortmann, The Theories of the Origin of the Antarctic Faunas and Floras, 

 American Naturalist, vol. 35, Feb. 1901, pp. 139-142. 



H. F. Osborn, Science, N. S., vol xi, April 13, 1900, pp. 564-566. 



2 K. Zittel [C. R. Eastman]. Text-book of Paleontology, vol. n, London, 1902, 

 pp. 179-187, translated by Eastman. See Notes 4, 5, pp. 575, 576. 



