620 PATAGONIAN EXPEDITIONS : ZOOLOGY. 



time. Such a hypothesis finds support in the presence of genera of the 

 old Brazilian type on the island of Fernando Noronha. Numerous .very 

 peculiar land-shell genera, such as those tabulated in fig. 25, have a dis- 

 tribution not readily explicable on any other hypothesis; while in still 

 other cases, allied but generically distinct groups are similarly distributed. 

 It may be noted that Bates has remarked that the Para insect fauna is 

 essentially Guianian.^ 



Figs. 26, 27 show the distribution of several old land-snail genera of 

 the Brazilian center. Fig. 25 that of several genera common to the 

 Brazilian and Guianian centers. 



That the Amazonian valley ever formed an upper Cretaceous strait 

 connecting the south Atlantic and Pacific, as claimed by Dr. Ortmann,^ 

 seems rather improbable. 



The Guiana-Colombian elevation has been a secondary radiation center for 

 a number of genera of autochthonous South American families, chiefly the 

 Biiliumlidcc. The arboreal groups Oxystyla and Corona have spread south 

 of the Amazon into eastern Brazil (fig. 28), while numerous other genera 

 from this center are restricted to the north and west as in fig. 29. The 

 Guiana-Colombian area also served as a secondary center for Antillean 

 and Mexican groups, entering by way of the Caribbean elevation and that 

 in the Panamic region. These groups have spread southward as in 



figs. 30-34- 



A hypothesis has been advanced by Dr. Ortmann^ that Archhelenis of 

 the Lower Cretaceous was succeeded in the Upper Cretaceous by a land 

 bridge from tropical Africa to an area covering Guiana, the Caribbean Sea 

 and the Mexico-Antillean region (the so-called Mesozonia), separated from 

 the Brazilian island. By this hypothesis, Antillia should be as rich in 

 African or Archhelenic types as Brazil, and in fact should show a closer 

 resemblance to the African fauna due to the later connection. This is 



' Naturalist on the Amazons, I, p. 109. 



^ Proc. Amer. Philos. Soc, XLI, p. 381, and in later articles. Dr. Ortmann's palseogeographic 

 maps incline strongly towards what Fiske would call the " wet theory." It is not likely that all 

 beds reported as Upper Cretaceous were below the sea at any one time. To map an Upper Cre- 

 taceous epicontinental sea to include all the exposures of a formation which included so long a 

 period of time is not warranted by our present slight knowledge of the stages of the South 

 American Cretaceous. 



' The Geographic Distribution of Freshwater Decapods and its Bearing upon Ancient Geography. 

 Proc. American Philosophical Society, 1902, pp. 380, 381. 



