268 THE WORLD OF LIFE 



and Australia in Early or Middle Tertiary times. This, however, 

 seems to me to be highly improbable for reasons given at full in my 

 Island Life. Its supposed necessity depends on the assumption 

 that the geological record is fairly complete, even as regards these 

 small mammals, and that their not being yet discovered in the 

 northern continents proves that they never existed there. But the 

 extreme rarity of the small Secondary Mammalia, though they have 

 been found scattered over the whole northern hemisphere, and the 

 limited area in South America in which these Tertiary marsupials 

 have been found, taken in connection with the enormous areas of 

 geologically unexplored land in Asia and Australia, should make us 

 very cautious in assuming such vast and physically improbable 

 changes of land and sea at such a comparatively recent epoch. 

 The theory of land-connection also introduces enormous difficulties 

 of various kinds which it is well briefly to consider. If we suppose 

 an absolute land-connection in order to allow the marsupial type to 

 have entered Australia from temperate South America, we have 

 to face the incredible fact, that of the whole varied mammalian 

 fauna of the latter country this one group only was transmitted. 

 In these same deposits there are found ancestral hoofed animals of 

 small size (Pyrotherium) ; numerous rodents allied to cavies and 

 porcupines; a host of Edentata allied to sloths, ant-eaters, and 

 armadillos. These, taken altogether, are many times more numer- 

 ous than the marsupials; they were more varied in structure and 

 mode of life; and it is almost incredible that not one representa- 

 tive of these somewhat higher forms sliould have reached the new 

 country, or having reached it should have all died out, while the 

 inferior group alone survived. Then, again, we know that birds 

 and insects must have abounded in South America at the same 

 period, while the whole 7000 miles of connecting land must have 

 been well clothed with vegetation to support the varied life that 

 must have existed upon it during the period of immigration. Yet 

 no indication of a direct transference or interchange of these nu- 

 merous forms of life in any adequate amount is found in either 

 Australia or South Temperate America. We can hardly suppose 

 such an enormous extent of land to have been raised above the 

 ocean; that it should have become sufficiently stocked with life to 

 serve as a bridge (7000 miles long!), and that a few very small 

 marsupials only should have crossed it; that it then sank as rap- 



