VARIATION IN TIME 15 1 



plication in the lines of suture, ignoring the general 

 form of the shell, its style of ornamentation,^ the 

 arrangement of the mouth, etc. We do not trouble 

 overmuch, moreover, about the chronological order of 

 the appearance of the fossil forms which one puts in 

 series. Thus to establish the pedigree of the Ursidae, 

 Gaudry and Boule introduce between the Hysen- 

 arctos of the upper Miocene and the first Ursi of the 

 Pliocene an actual genus, the (Eluropos of China, 

 because this animal realizes, from the point of view 

 of the progressive development of the tuberculous 

 teeth, an intermediate state between the two 

 genera it is sought to connect. These are abso- 

 lute anachronisms, and, to my mind, utterly in- 

 admissible. 



This method assuredly presents the greatest 

 dangers, because it leads one to confound with the 

 real evolution of a group what is, in fact, simply 

 the functional evolution of an organ in a series of 

 genera belonging to different natural branches 

 having no kind of ancestral kinship between them. 

 It could not fail to lead to the manufacture of arti- 

 ficial and incorrect concatenations, as I have already 

 had occasion to point out regarding the families of 

 the Horses and the Bears. It may not be useless to 

 give a few more examples. The evolution of the 

 family of the Rhinoceroses has been studied by 

 Gaudry from the gradual development of the 

 nasal bones since the Palceotheria of Eocene times 

 down to the two-horned rhinoceros of the present 

 day. The series commences with the Palceo- 

 therium medium, whose slender and short nasal 

 bones suggest the existence of a fleshy trunk ; 



