MAMMALIA. 247 



and if these are not the direct ancestors of the warm-blooded 

 quadrupeds termed mammals, they must at least be extremely 

 close allies of those ancestral forms. Unfortunately, however, 

 no fossils of this character have hitherto been found in Mesozoic 

 strata of later date than the Trias in any part of the world. 

 It is thus still impossible to trace the evolution upwards to 

 the base of the Caiuozoic or Tertiary strata, in which typical 

 and well-preserved mammalian skeletons occur in abundance. 

 The only mammalian fossils hitherto discovered in the Jurassic 

 and Cretaceous rocks are merely fragments of a most unsatis- 

 factory nature. They represent tribes of dwarfed animals not 

 larger than rats, some probably related to the existing mono- 

 tremes, others to the marsupials, none clearly belonging to the 

 highest mammalian type (that of the Eutheria) or linking this 

 with the grades below. 



It seems, however, certain that most of the Mesozoic 

 Mammalia, if not all, were of the primitive types which at the 

 present day are either oviparous (Prototherian or Ornitho- 

 delphian) or non-placental (Metatherian or Didelphian); for 

 nearly all the earliest Tertiary skeletons exhibit some features 

 now peculiar to the latter, and these fossils undoubtedly 

 represent the generalized forerunners of the typical placental 

 mammals (Eutheria) which are dominant throughout the 

 Tertiary period. The gradual evolution of the principal 

 Eutherian orders and their smaller divisions can be traced by 

 comparing the skeletons from the successive Tertiary forma- 

 tions ; and it will be noticed that the chief modifications 

 occur in the limbs and teeth, the parts by which they are 

 brought into most intimate relationship with their environ- 

 ment. They seem to have inherited a generalized, plantigrade, 

 five-fingered hand and foot from their cold-blooded ancestors, 

 and these extremities soon undergo strange modifications never 

 observed in any type below the mammalian. They are also 

 assumed to have inherited simple conical or triconodont teeth 

 (though there is as yet no demonstration of this), and these 

 teeth immediately become complicated in various ways, ap- 

 parently in accordance with certain well-defined laws of folding 

 and cusp-development. 



