KOSS ON EVOLUTION. 425 



Fox-like — these last being somewhat carnivorous in their habits. 

 The remains of Monkeys have been found in the Eocene in 

 America, and these are found to be of less differentiated types than 

 existing American Monkeys, and have characteristics which ally 

 them to the existing forms of the next lower grades of the Mam- 

 malia, the Carnivora and the JJngulata or Herhivora, and it is a 

 curious and instructive fact that each of these Orders was at first, 

 without exception, plantigrade, that is, walked on the entire foot as 

 does man and do Monkeys, so that the later digitigrade types were 

 reached in each case by a gradual differentiation. Insectivores, 

 Rodents, and Marsupials present a similar series of types, the lower 

 being plantigrade and the higher, and later, digitigrade. All the 

 Edentata and Monotremata are plantigrade. All the Mammalian 

 remains of the Eocene are of highly generalized types. 



Didelphia consists of Series differentiated so similarly to those 

 oi Monodelphia. as to have the same names applied to them, viz : 

 Marsupial Monkeys, Carnivores, Herbivores, Insectivores, and 

 Rodents, That Didelphia had at the time of its greatest expan- 

 sion marine representatives, corresponding to Cetacea and Sirenia 

 among the Monodelphia is altogether probable. A somewhat 

 similar parallelism exists between the various Orders of Reptiles 

 and Amphibians. Indeed as we trace each great group forwards 

 in time we find a constantly progressive differentiation outward 

 from the general to the particular, or special, from the ominivorous 

 for example, to the more and more completely herbivorous, carni- 

 vorous, or insectivorous, and from these to others having still more 

 specialized habits as to food, and all the corresponding peculiarities 

 of organization and instinct. 



Again there is the tendency outwards as to habitat — to occupy 

 the land, the water, or partly each of these, and that in every 

 climate. Each of these differentiate into flying and non-flying, and 

 some of each of these into climbers, and some into burrowers * 

 in fact each subordinate group as it expands has a tendency to 

 repeat from its ow^n starting point all of these differentiations, and 

 a thousand minor ones ; so that each of these differentiations may 

 be more or less fundamental than other co-existing ones. Thus in 

 the Chieroptera the adaptability for flight seems more fundamental 



