GEOGRAPHICAL DISTRIBUTION AND DESCENT. 



i;7 



number of types in common, a fact most 

 probably to be explained by the supposition 

 that the types in question have radiated in 

 different directions from the extreme north, 

 that portion of the earth having had, even in 

 the Miocene period, a mild climate similar to 

 that of Southern Europe at the present day. 



North America is 

 only slightly differ- 

 ent from the region < 

 just spoken of. Out 

 of its twenty-four 

 genera only eleven 

 belong to it alone; 

 the others are shared 

 by it with South 

 America, or with the 

 Eurasian continent. 

 The types peculiar 

 to it, however, are 

 so sharply special- 

 ized that we may 

 regard them as de- 

 scendants of separ- 

 ate stocks. 



It may justly ex- 

 cite surprise that 

 India and the Sunda 

 Islands are so poor 

 in rodents. This 

 region, which is so rich in the other mam- 

 mals, is so sparingly supplied with rodents 

 that we may well suppose that all the forms 

 represented there are due to immigration. Of 

 sixteen genera known to exist in India and 

 the Sunda Islands only four are peculiar to 

 these regions, three of the four belonging to 

 the mouse family, and one to that of the 

 porcupines; and there is even something 

 arbitrary in the assigning of generic charac- 

 ters to the four genera in question. If the 

 East Indies have in the case of the other 

 orders of mammals contributed many of the 

 primitive stocks, in the case of the rodents it 

 has furnished not one characteristic form ; the 

 few types represented there are derived from 



Vol. n. 



Fig. 239, — The Rabbit {Lefus ciinicidus). page 175. 



Other regions. Now this remarkable result 

 can be due neither to the influence of climate 

 nor to deficiency of food, for South America 

 and some parts of Africa, which have the 

 same conditions of existence, are richly sup- 

 plied with rodents. 



We are met with similar peculiar phe- 

 nomena in studying 

 the distribution of 

 the families. We 

 have already on 

 several occasions di- 

 rected attention to 

 the fact that the sub- 

 division of the ro- 

 dents into families 

 and genera is pecu- 

 liarly difficult, and 

 that investigators 

 still hold contrary 

 opinions as regards 

 this matter. But 

 whatever view one 

 may adopt, one is 

 bound to admit that 

 certain types are re- 

 markably widely dis- 

 tributed, while others 

 are limited to more 

 or less restricted 

 areas. As regards the latter, South America 

 exhibits the greatest amount of peculiarity, 

 for, apart from the northern regions, the two 

 hemispheres are sharply separated from one 

 another, and illustrate that contrast that has 

 often been dwelt on. 



There is in fact omy a single family dis- 

 tributed over the whole earth, including 

 Australia and Madagascar. This is the re- 

 markably numerous family of the mice, which 

 has adapted itself to all climates. And then 

 there are two others, the hares and squirrels, 

 which are excluded only from the regions 

 mentioned. To the Old World exclusively 

 belong only the mole-rats, the dormice, and 

 the typical porcupines; the beavers, the 



65 



