50 PARALLELISM IN DEVELOPMENT. 



ruminants is, however, eomixiratively remote, but in tlie 

 latter group and tlie camels tliese teetli are so alike as to 

 require an esx^ert to distinguish between them. Never- 

 theless, there is good evidence to show that the camels and 

 the ruminants, if not also the chevrotains, have acquired 

 their crescent-like molar teetli Cjuite independently of one 

 another, and it therefore jet remains for those writers 

 who explain evolution bj some mode of what they are 

 pleased to call natural selection to account adequately for 

 the similarity thus existing between structures of such 

 totally different origin, when they could have been made 

 equally efficient if unlike. 



Passing from the consideration of teeth to that of limbs, 

 we may mention the remarkable similarity displayed in the 

 mode by which the lower segments of the limbs of the 

 even-toed and odd -toed hoofed mammals have been gradu- 

 ally elongated by the formation of a cannon-bone and the 

 disappearance of either three or four of the lateral digits ; 

 the cannon-bone in the horse consisting of but a single 

 element carrying one digit, while in the ruminants it 

 comprises two united elements supporting a pair of toes. 

 This IS clearly a case of ^parallelism in develo^jment 

 attained by a slightly different modification of jsrinciple. 

 The ]:>arallelisin does not, however, stop here, since an 

 essi'utially similar type of cannon-bone has been jjroduced 

 in Ijirds ; only that in that group (with the exception of the 

 ostrich) three long bones enter into its composition, which 

 is further complicated by the addition of a bone from 

 the ankle above. Seeing that in all these groups the 

 l^arallelism has been arrived at by a different structural 

 modification, the explanation of its mode of evolution is 

 much less difficult than in the ease of the molars of the 

 camels and ruminants, where, as we have seen, the struc- 

 ture is practically identical. 



Recent discoveries in North America have brought to 

 light the existence of a kind of secondary parallelism 

 among certain ])eculiar mammals which ma)^ be included 

 among the hoofed or ungulate division of that class. In 

 the even-toed group of that division, as exemplified by 

 the pigs and ruminants, it is the third and fourth dibit's 



