162 MAMMALIAN DESCENT. [Legt. VI. 



Shrews, l)iit it must li;i\'e taken them more time than 

 we can conceive of to settle down in their colonies, and 

 to people the greater part of the globe with their pigmy 

 tribes. This is not brought forward as though it were 

 an insuperable difficulty — as a problem that cannot be 

 accounted for on evolutional principles. What was 

 wanted was an abundance of time. The forefathers of 

 the Shrews were certainly early Eutheria that became 

 modified mainly hj dwarfing, and so did not develop 

 upwards into noljler types. 



Amongst other kinds of early Eutheria an opposite 

 kind of variation took place. They became larger and 

 larger, and according to the difference of their feeding- 

 grounds, these being sometimes flat or hilly, soft or 

 hard ; on burning sands, or on frozen snow ; their feet 

 became modified, the innermost digit going first. The 

 tooth-pattern, also, grew more and more complex, as the 

 diet Ijecame in many cases of a purely vegetable kind. 

 Then a purely vegetable diet is of necessity very varied ; 

 browsing is added to grazing; and thus the molar teeth 

 have to be very large, with large bones to hold them, and 

 the grinding crowns must he very effective mill- 

 stones. 



I may here give an instance of the freedom of 

 selection as to the kind of food eaten by the most 

 purely herbivorous ty})e. The Horse, every one knows, 

 is a most particular creature as to his diet ; his sensitive- 

 ness to the odours of various kinds of food, especially, is 



