i6 



JUK MAMMALIA. 



Miocene orders and families are there unknown; 

 the monke)'s, edentates, ruminants, and carnivores, 

 so widel)' diffused on the neiglibouring mainland, 

 are all alike absent here; the bats and rodents, 

 which arc found on these islands, belong to an 

 ancient Eocene nobility, and so also do the insecti- 

 vorcs, which are wholly wanting on the mainland of 

 South America. 



These two regions liave followed a separate course 

 of development, at least since the Eocene period. ' 



And here now wc are met by another fact. If the 

 Eocene and Miocene genera found on the two 

 hemispheres and described by paLxontologists arc 

 compared, we find with astonishment that, with the 

 exception of two genera belonging to the Perisso- 

 dact)'la, the genera Coryphodon and Lophiothcrium, 

 derived from the earliest Eocene strata, there are 

 none common to both hemispheres. What con- 

 clusion must be deduced from this fact.'' Assuredly 

 none but this, that the two great divisions of the 

 earth were already separate in those times. But 

 there are in both corresponding series in the 

 evolution of mammalian forms; monkeys, horses, 

 ruminants, and carnivores exhibit parallel series on 

 the two sides of the ocean. 



In the Pliocene and Quaternary periods the two 

 hemis|)heres are not so sharply distinguished in the 

 north. In the Quaternary, in which the polar 

 species of the two hemispheres advanced in general 

 further to the south, there is not only, as at present, 

 similarity but entire agreement. These facts made 

 the adoption of a separate circumpolar region 

 necessary. We cannot associate this region with 

 those immediately to the south, when we find that 

 the latter are not only absolutel)' distinct on the 

 two sides of the ocean at the present day, but 

 were equally distinct to the northern shores of the 

 respective continents down to a period geologically 

 recent. 



The two parts of the mainland of America are 

 connected by the Isthmus of Panama, b)' means of 

 which many an interchange has taken place. Ikit 

 on the whole the two land-masses referred to 

 are astonishingly different. North America, llie 

 Xearctic region, has neither monkeys, nor manatees, 

 nor tapirs, camels, agoutis, or edentates, like South 

 America, the Neotrofjical region, which again has 

 neither inscctivores, nor hollow-horned ruminants 

 or pouched rats. These two regions are accord- 

 ingly sharply marked off from one another. 



Quite different is it, however, with the three 

 regions of the Old World. The Pal.earctic region has 



neither edentates, pi'osimians, nor elephants, whii,h 

 the other two regions possess; but for the most 

 part it is necessary to descend to families and even 

 to genera iri order to make out general distinctions 

 between the three regions. We must also grant 

 that their boundaries become obliterated when wc 

 go back from the present time to earlier periods. 

 The three regions finally merge entirely into one. 

 In the Miocene and Ivjcene jicriods the Palrearctic 

 region had edentates, giraffes, and prosimians, and 

 as late as the Quaternary period elephants, hippo- 

 potamuses, rhinoceroses, and hyaenas. The present 

 limits of the region have accordingly only gradually 

 been established. 



If we now go on to consider the classification 

 which may be deduced from and ought to sum up 

 all the results of biological, developmental, and 

 geographical investigations, we must confess that 

 considerable impediments stand in the way. T'lrst 

 of all we recognize the pierfectly clear and sharp 

 distinction between aplacental and placental mam- 

 mals, but after we have got so far we find the 

 principles of a further subdivision, especially of the 

 placental mammals, difficult to determine. 



The Eocene orders furnish us with nailed and 

 hoofed forms, Deciduata and Indeciduata, forms 

 with diffuse, zonary, and discoidal placentas, with 

 complete and incomplete dentition. All these above- 

 named orders cannot be derived from one another. 

 If, as is very probable, the\' are all descended fiom 

 marsupial stocks, the)' must trace back their descent 

 along several lines of different origin. Let us take 

 for granted for an instant this descent from a 

 marsupial ancestr)-, and see what conclusions may 

 be deduced therefrom. Such conclusions can be 

 based onl)- on the dental S3-stem, which is all that 

 is known to us in the case of the earliest mammals. 



Presumable Relations of existing Placental 

 Orders to a Marsupial Ancestry as shown by 

 the Dental Characters. The insectivorous type 

 is the oldest. We have seen that the marsupials 

 of the Trias and Oolitic (Jura) periods are insccti- 

 vores. This is accordingly a very old, perhaps 

 e\en the ])rimiti\c stock'. 



The bats aie onl)' flying inscctivores. If we 

 unite the inscctivores with webbed feet and those 

 with the toes free, we might also, without being 

 guilty of an)' great offence, add to the same group 

 those with flying membranes. 



Some Prosimii are manifestly inscctivores. They 

 are probably descended accordingly from insecti- 

 vorous marsupials. Others, however, likewise very 



