IN THE TAIL. 



63 



from the upper jaws, wliicli were totally devoid of front 

 or incisor teeth. Had such skulls been discovered without 

 any indication as to the nature of the limbs with which 

 they were associated, they would inevitaldy have been 

 assigned to the same group of animals. The resemblance 

 existing between them, is, however, clearly due to parallel 

 development, and we are thus shown another striking 

 instance of caution necessary in endeavouring to determine 

 the aihuities of extinct animals from the evidence of in- 

 comjjlete remains. 



Fig. 19.— StuU of Protoceras. (After Osborn.) 



Our last instance of parallelism is alluded to in a later 

 chajjter, entitled " 'ihe Oldest Fishes and their Fins," in 

 the course of which it is shown that while both the most 

 ancient birds and the oldest fishes had long tapering tails 

 with the joints of the Viackbone gradually diminishing in 

 size, and each carrying either a jjair of feathers or a pair 

 of fin-rays, in all the modern representatives of the former 

 grouj), and in a large section of the latter, the end of the 

 vertebral column has been aborted into a composite bone, 

 from which either the feathers or the rays of the tail 

 diverge in a fan-like manner, 



In conclusion, we may say that although there is no 

 very great difficulty in satisfactorily a(;counting for external 

 parallelism obviously due to the necessity for adaptation 

 to a particular mode of life, or in expilaining those instances 

 where a particular result has been brought about by 



