32 ANNUAL OF SCIENTIFIC DISCOVERY. 



Geology gives a wider range to the horse and elephant kinds than was cog- 

 nizant to the student of living species only. The existing Equidae and Ele- 

 phantidse properly belong, or are limited to, the Old World; and the 

 elephants to Asia and Africa, the species of the two continents being quite 

 distinct. The horse, as Btiffon remarked, carried terror to the eye of the 

 indigenous Americans, viewing the animal for the first time, as it proudly 

 bore their Spanish conqueror. But a species of Equus, coexisted with the 

 Megatherium and Megalonyx, in both South and North America, and per- 

 ished apparently with them, before the human period. Elephants are 

 dependent chiefly upon trees for food. One species now finds conditions of 

 existence in the rich forests of tropical Asia; and a second species in those 

 of tropical Africa. Why, we may ask, should not a third be living at the 

 expense of the still more luxuriant vegetation watered by the Oronooko, the 

 Essequibo, the Amazon, and the La Plata, in tropical America? Geology tells 

 us that at least two kinds of elephant (Mastodon Andium and M. Humboldtii) 

 formerly did derive their subsistence, along with the great Megathcrioid 

 beasts, from that abundant source. We may infer that the general growth 

 of large forests, and the absence of deadly enemies, were the main conditions 

 of the former existence of elephantine animals over every part of the globe. 

 We have the most pregnant proof of the importance of Palaeontology in 

 rectifying and expanding ideas deduced from recent Zoology of the geograph- 

 ical limits of particular forms of animals, by the results of its application to 

 the proboscidian or elephantine family. But such retrospective views of life 

 in remote periods, in many important instances, confirm the Zoologist's 

 deductions of the originally restricted range of particular forms of mamma- 

 lian life. The sum of all the evidence from the fossil world in Australia 

 proves its mammalian population to have been essentially the same in pleis- 

 tocene, if not pliocene times, as now; only represented, as the Edentate 

 mammals in South America were then represented, l>y more numerous gen- 

 era, and much more gigantic species, than now exist. But Geology has 

 revealed more important and unexpected facts relative to the marsupial type 

 of quadrupeds. In the miocene and eocene tertiary deposits, marsupial fos- 

 sils of the American genus Didelphys have been found, both in France and 

 England ; and they are associated with Tapirs like that of America. In a 

 more ancient geological period remains of marsupials, some insectivorous, as 

 Spalacotherium and Triconodon, others with teeth like the peculiar premo- 

 lars in the Australian genus Hypsipromnus, have been found in the upper 

 oolite of the Isle of Purbeck. In the lower oolite at Stonesfield, Oxfordshire, 

 marsupial remains have been found having their nearest living representa- 

 tives in the Australian genera Myrmecobius and Dasyurus. Thus it would 

 seem, that the deeper we penetrate the earth, or, in other words, the further 

 we recede in time, the more completely are we absolved from the present 

 l:nvs of geographical distribution. In comparing the mammalian fossils 

 found in British pleistocene and pliocene beds, we have often to travel to Asia 

 or Africa for their homologues. In the miocene and eocene strata some 

 fossils occur which compel us to go to America for the nearest representa- 

 tives. To match the mammalian remains from the English oolitic formations, 

 we must bring species from the Antipodes. These are truly most suggestive 

 facts. If the present laAvs of geographical distribution depend, in an impor- 

 tant degree, upon the present configuration and position of continents and 

 islands, what a total change in the geographical character of the earth's sur- 

 face must have taken place since the " Stonesfield slate " was deposited in 



