THE GEOLOGICAL RECORD 301 



three main subclasses. (1) Prototherid, with the single surviving 

 order Monotremata, represented by the Australian spiny ant- 

 eater (Echidna, Fig. 92) and duck-billed Platypus (Ornithorhyn- 

 chus, Fig. 91), which, as we have already seen, are intermediate 

 both in anatomical structure and in their method of reproduction 

 between the reptiles and the more typical mammals. (2) Meta- 

 theria, with the sole order Marsupialia, represented to-day by 

 the Australian kangaroos, wombats, phalangers, thylacines and 

 numerous other pouched mammals, and by the American opossums 

 (Didelphyidae) and Csenolestes. These are in some respects primi- 

 tive forms, in which the young are born at a very early stage of 

 their development (there being at the most only a very feebly 

 developed placenta) and usually carried about, attached to the 

 teat of the mother, in a marsupium or pouch. They are readily 

 distinguished by osteological features, especially the structure of the 

 juws and teeth (Fig. 108), from the higher mammals. (3) Eutheria, 

 the dominant mammals of the present day, with a large number 

 of orders edentates, whales, sirenians, ungulates, rodents, 

 carnivores, insectivores, bats and primates ; with well developed 

 placenta and retaining the young in the womb until a very 

 advanced stage of development. There can be no doubt that 

 the Eutheria are on the whole more highly organized forms 

 than the Metatheria and the latter than the Prototheria. This 

 is especially indicated by the gradual acquisition and perfection 

 of the characteristically mammalian method of nourishing the 

 young by means of the placenta and mammary glands. 



The order of appearance in time of the principal subdivisions 

 of the class Mammalia, as indicated by the geological record, 

 entirely supports the evolutionary hypothesis. The earliest 

 known fossils which may possibly be referable to the class are of 

 Triassic age. Tritylodon is represented by an imperfect skull 

 from the Karoo formation of South Africa. This has generally 

 been regarded as mammalian, but Dr. Smith Woodward, notwith- 

 standing the fact that it has tuberculated grinding teeth with 

 deeply cleft roots, considers that it probably belonged to an 

 anomodont or theromorph reptile. Microlestes is represented by 

 double-rooted, multituberculate teeth obtained from European 

 Rhaetic formations, and may also possibly be reptilian. 



In deposits of Lower and Upper Jurassic age we find more 

 convincing evidence of the existence of true mammals. The 

 celebrated Stonesfield slate of Oxfordshire (Lower Jurassic) has 



