CONVERGENT EVOLUTION 253 



three (i.i i. 3 ) in the lower jaw ; and one canine (c.), three 

 premolars (p.m.i ) and four molars (m.i 4) in each jaw. The 

 thylacine skull is further distinguished from that of the dog by 

 the presence of large vacuities in the hinder part of the bony 

 palate and by the strongly marked inflection of the angles of the 

 mandibles. These two characters, both of which are shown in 

 the illustrations, as well as the arrangement of the cheek teeth 

 and other minor features which it is unnecessary to specify, are 

 fundamental peculiarities of the great group Marsupialia and at 

 once indicate the true affinities of Thylacinus. 



The case of the Australian marsupial mole, Notoryctes, is 

 equally striking. The powerfully built digging limbs, the soft, 

 close fur, the absence of external ears and the loss of sight, as 

 well as the general shape of the body, all in adaptation to its 

 burrowing habits, cause it to assume a wonderfully mole-like 

 aspect, while in reality it comes nowhere near the true moles as 

 regards genetic relationship. 



Perhaps the most remarkable of all known cases of convergent 

 evolution, however, is met with in the Ungulata. We have 

 already seen that amongst the typical ungulate mammals of the 

 present day we can distinguish two series, odd-toed or perisso- 

 dactyl, and even-toed or artiodactyl. In the odd-toed series the 

 reduction of the digits culminates in the horse, with a single 

 perfect digit in each foot and vestiges of two others in the form 

 of splint-bones. This extreme modification of the unguligrade 

 type is so peculiar that it is difficult to believe that precisely the 

 same line of evolution has been followed independently by two 

 different groups of animals. Yet such appears to be actually 

 the case. There is an extinct group of ungulates known as the 

 Litopterna, whose remains have been described by Ameghino 

 from Tertiary beds of Patagonia. The small size of the 

 brain-cavity, the characters of the dentition, cervical vertebrae, 

 carpus and tarsus, indicate that they were more primitive forms 

 than the true Perissodactyla, and, as Dr. Smith Woodward 

 observes, they reached their maximum of specialization at an 

 earlier period than the latter. We find amongst them forms 

 (Theosodon, Fig. 109, A, B) with three well-developed toes as in 

 the rhinoceros, forms (Proterotherium, Fig. 109, C) with one well- 

 developed toe and two small ones as in some of the extinct 

 ancestors of the horse, and forms (Thoatherium, Fig. 109, D, E) 

 with a single toe as in existing horses ; and in all cases the axis 



