172 THE TRANSFORMATIONS OF THE ANIMAL WORLD 



jaw) for Osborn not to hesitate to consider it as 

 the ancestral form of our small carnivorous Marsu- 

 pials, such as the existing Didelphs or Sarigues. 

 We here see, then, a first branch of lower Mammals 

 which extends, with few interruptions and geo- 

 graphical displacements, from our present period 

 back to the Trias, and probably very much further. 

 The order of the Multituberculata supplies us 

 with a second branch of quite as great geological 

 longevity. We have from the Rhaetian of Wurtem- 

 burg and England the isolated molars of a quite 

 small Mammal, the Microlestes, which, notwith- 

 standing the paucity of proofs, really seems to be 

 the most ancient known representative of the 

 family of Plagiaulacidce. The genus Plagiaulax, 

 the type of the family, is also only known by some 

 small mandibles furnished with a large conical in- 

 cisor analogous to that of the Rodents, with four 

 compressed molars, sharp at the top, and orna- 

 mented with oblique lines on the sides, and, lastly, 

 with two little cup-like posterior molars surrounded 

 by five small tubercules. The Plagiaulax has been 

 found in the strata of Purbeck, that is to say, in 

 the last layers of the Jurassic, and, notwithstanding 

 the great geological lacuna which separates it from 

 the Microlestes, no palaeontologist disputes the 

 close affiliation of these two genera. After another 

 no less important lacuna, which comprises nearly the 

 entire Cretacean, we find in the terminal strata of 

 the American Cretacean (Laramie stage) a new 

 representative of this family, the Ctenacodon, and, 

 a little later, in the lower Eocene of Cernay near 

 Rheims, another form so closely allied to the 



