INTRODUCTION. 



Can we fill up the gap by supposing that there 

 arose at some period within it a single placental 

 type from which all the present forms are descended? 

 Or must we assume more than one primitive 

 placental stock to account for the present diverg- 

 ences among the members of the class? The 

 answers to these questions must necessarily be 

 based largely on speculation, and though the facts 

 that have been ascertained may justify pretty 

 confident conclusions with regard to some of them, 

 we are inevitably restricted to hypothesis when we 

 attempt to trace the history or evolution of any 

 particular group. 



So much, however, is certain. After the long 

 gap between Jurassic and Tertiary times there 

 appear with the Eocene both in the Old World 

 and the New the following still surviving orders: 

 Prosimii, Insectivora, Carnivora, Perissodactyla, 

 Artiodactyla, Rodentia, and in the Old World alone, 

 the Chiroptera. All these orders are represented 

 either by families which are now quite extinct, or 

 by equivocal types which cannot be referred to any 

 family, or, lastly, by forms still existing. Thus the 

 Prosimii in France have furnished a genus allied to 

 the Potto, while the other families are extinct. 

 Further, leaf-nosed and smooth-nosed forms are 

 found among the bats; tanrecs, moles, and shrews 

 among the Insectivora; Canida and Viverrida 

 among the Carnivora; tapirs and horses among 

 the Perissodactyla; pigs among the Artiodactyla; 

 squirrels, dormice, mice, degus (genus Octodon), 

 and spiny rats (Echimyida) among the rodents, 

 together with extinct families and indeterminate 

 types, from which families better characterized can 

 be derived. 



This accordingly is the oldest nobility among 

 the placental mammals, whose family trees can be 

 traced back to the Eocene. The roots manifestly 

 reach much further back, but we have no certain 

 knowledge of them. 



Let us now consider this fact in the light of the 

 present distribution of mammalian forms. Seeing 

 that all these orders in families already existed in 

 Eocene times, if there is any region of the globe 

 from which they are altogether absent, or in which 

 they are poorly represented, that deficiency must 

 be accounted for. It is universally found that when 

 a higher type appears it tends to displace the lower. 

 The latter becomes extinct, or gets reduced to a 

 few representatives enabled by special circumstances 

 or habits to survive. They may escape even amidst 

 the competition of higher forms through the fact of 



their living on trees, and thus being out of the 

 reach of many enemies, or through being able to 

 hide in burrows, or being nocturnal in their mode 

 of life. But, apart from such special favouring 

 conditions, the doom of the lower type is to dis- 

 appear before the higher, and since the higher 

 placental forms were already largely represented 

 in Eocene times, we must ask how it is that any 

 region of the earth has escaped being overrun by 

 them. 



Such a region does exist in Australia and some 

 of the neighbouring islands. The fauna of that 

 region comprises but few placental forms among 

 its mammals, but a great number and variety of 

 such as have no placenta. There is only one way 

 of accounting for such a fact as this. Before the 

 appearance of the chief representatives of the 

 placental type, that is to say, before the Eocene 

 epoch, the Australian region must have been cut 

 off from the other lands of the globe ; it must thus 

 have been saved from that struggle for existence 

 in which the higher forms of the Eocene epoch 

 took part. 



Passing on now to the Miocene we find that in 

 that epoch the number of the orders is completed. 

 None of the orders of the present day is wanting 

 in Miocene formations. Monkeys, seals, whales, 

 sea-cows, elephants, and edentates all appear. The 

 families become specialized, the monkeys of the 

 Old and New World, hedgehogs, felines, hyaenas, 

 martensand bears, rhinocerosesand hippopotamuses, 

 as well as ruminants, can be more and more clearly 

 discriminated. In short we see how in the strata of 

 the Miocene the types follow one another in higher 

 and higher degrees of specialization. 



It is clear, moreover, that those regions which 

 present to our view only the orders appearing 

 first in the Eocene, and which are without the 

 orders originating in the Miocene, were isolated 

 before the latter epoch, though not till after the 

 Eocene. 



In this position are Madagascar and the Antilles. 

 The great African island is mainly inhabited by 

 prosimians; but there are found in addition bats, 

 insectivores, rodents, a species of pig, viverrines, 

 and the old type Cryptoprocta. No vestiges, how- 

 ever, have yet been found there of any order or 

 family belonging to the Miocene neither monkeys, 

 nor ruminants, nor edentates, which are all so 

 abundant, nevertheless, both on the mainland of 

 Africa and in India. The Antilles present an 

 analogous case, differing only in the details. The 



