214 



THE MARSUPIALS. 



the teeth of Stereognathus and those of the 

 still living koala. 



It is important to observe further that the 

 hoofed mammals, at least as regards their 

 dentition, are descended from an animal be- 

 longing to the Jurassic period whose marsu- 

 pial character is not indeed perfectly demon- 

 strated, but is nevertheless highly probable. 



A wide gap separates the marsupials of the 

 Stonesfield Slates from those of the "dirt- 

 bed " of the Purbeck strata. Geologists still 

 dispute, not indeed about the relative position, 

 but about the classification of these strata; if 

 they are referred to the Jurassic system then 

 they form its uppermost story; if, on the 

 other hand, they are referred to the Creta- 

 ceous system, then they form its base. But 

 whichever of these two systems they be as- 

 signed to, the fact remains, that between those 

 Purbeck strata and the lower members of the 

 Tertiary series there lies the whole extent of 

 the Cretaceous system, which has hitherto 

 appeared to be altogether devoid of mam- 

 malian remains. 



In the case of most of the jaws found in 

 the Purbeck region the marsupial character 

 is unequivocally demonstrated by the occur- 

 rence of the inflexed angle of the lower jaw; 

 in the case of the others it appears to be 

 placed beyond doubt by the great similarity 

 of the teeth to those of actual marsupials of 

 the present day. In general the insectivorous 

 type still prevails: Spalacotherium, Amblo- 

 therium, Peralestes, Stylodon, Balodon, are 

 characteristic insectivores ; Triconodon, Tria- 

 canthodon, already present more resemblance 

 to the carnivorous Tasmanian wolf, a living 

 marsupial form, as the reader will remember. 

 Only the genus Plagiaulax presents a quite 

 exceptional type. While Owen refers it to 

 the carnivorous mammals and ranks it with 

 the genus Thylacoleo of the Quaternary 

 strata of Australia, other investigators regard 

 Thylacoleo and Plagiaulax as herbivorous and 

 allied to the rat-kangaroos. In the strata of 

 the Upper Jura of Wybming in America, 



corresponding to the Purbeck Beds of Eng- 

 land, still other remains have been found, 

 some of which (Ctenacodon) are allied to 

 Plagiaulax, and others again (Dryolestes, 

 Tinodon) appear, on the contrary, rather to 

 be insectivorous in their relations. 



At the end of the wide interval of the 

 Cretaceous Period there at last appear in the 

 Upper Eocene and Lower Miocene of Europe, 

 besides Plagiaulax, the genus Peratherium, 

 allied to the opossums, and the genus Gale- 

 thylax, both of which may be regarded as 

 sub-genera of the present Didelphyida. The 

 smaller species of the latter still retain a per- 

 fect insectivorous dentition, while in the larger 

 species the shortening of the inner cusps or 

 peaks in the molars indicates a relationship to 

 the carnivorous teeth provided with a heel 

 such as is seen in the martens and Viverrida. 



With respect to the dentition we should 

 thus be able to establish the existence of an 

 almost unbroken chain of insectivorous mar- 

 supials stretching down from the earliest period 

 in which there are any mammalian remains 

 whatever to the present day, though in doing 

 so it would certainly be necessary to note the 

 tendency to a bunodont dentition "manifested 

 in the genus Stereognathus of the Stonesfield 

 Slates, and the tendency to a carnivorous 

 dentition that appears still later in the remains 

 found in the Purbeck strata. The dentition 

 of Plagiaulax might even be regarded as 

 representing a peculiar type, which would 

 lead up either to the rodents or to elephants. 



The different views that have been put 

 forward regarding the transitional forms ^f 

 dentition such as are represented by the 

 genera Stereognathus, Plagiaulax, and Thy- 

 lacoleo, show that these are indeterminate 

 equivocal types, from which different forms of 

 dentition with well-marked characters might 

 be developed; that, in other words, they are 

 collective types exhibiting a union of char- 

 acters which afterwards come to be separated. 



Another trait in the organization also 

 deserves to be noted. 



