148 THE ARCTOG^EIC REALM. [CHAP. 



the same family group, which are only known from strata of the 

 Oligocene and Miocene epochs, marsupials appear to have been 

 absent from a large part of the realm during the Tertiary period, 

 although there is reason to believe that during the Eocene they 

 must have survived in south-eastern Asia 1 . Among volant 

 mammals, bats of the Neogaeic family PhyllostomatidGR* are now 

 absent from the whole of the realm, with the exception of a part 

 of the Pacific side of North America. Again, to revert to the 

 non-volant forms, the Lemuroid suborder of the Primates seems to 

 have been absolutely restricted to Arctogaea, at least since the 

 Miocene epoch, although it may turn out that the monkeys and 

 marmosets of South America may be descended from ancestral 

 lemuroids which inhabited that country at an epoch previous 

 to the deposition of the Santa Crucian beds. 



To take a comprehensive survey of the whole Secondary and 



Communit Tertiary mammalian faunas of Arctogaea would entail 



of earliest such a mass of palaeontological and anatomical 



detail that it would only weary the majority of our 



readers, and we must accordingly limit ourselves to noticing some 



of the most striking features in the earlier faunas, and then pass 



on to the consideration of some of the more widely spread modern 



groups. 



In the chapter devoted to the Notogaeic realm, it has been 

 already pointed out (p. 51) that during the Jurassic period Europe 

 and North America were populated with a fauna of polyprotodont 

 marsupials of small size, some of which appear to have been the 

 ancestral types from which those now inhabiting the Notogseic 

 and Neogaeic realms were derived, while others have disappeared 

 entirely. It will be unnecessary to recapitulate the names of the 

 more representative of these forms, but it may be stated that while 

 the fauna of the lower Jurassic Stonesfield Slate has no equivalent 

 in North America, that of the upper Jurassic Purbeck beds of Dor- 

 setshire is paralleled in the latter area. Although some difference 

 of opinion prevails among palaeontologists as to the identity of the 

 American with the European genera, there can be no doubt that 



1 Vide suprd, p. 55. 



2 The genus Necromantis, from the French Oligocene, has been assigned 

 to this family. 



