A MOUNTAIN TARN, 173 



small colony in the tarn, which would thus 

 have room to develop in their own way to 

 suit their own peculiar circumstances, without 

 any cross of fresh blood from the old stock 

 to keep them true to the general type of the 

 race in the lowland rivers. For, as Mr. A. 

 R. Wallace has ingeniously pointed out, such 

 isolated mountain pools are really the aquatic 

 equivalents or analogues of oceanic islands. 

 In all such limited and hermetically closed 

 habitats every stray denizen is liable very 

 rapidly to undergo considerable changes. 

 The nature of the food-stuffs is new, and 

 their variety but scianty ; the enemies, if any, 

 are fewer in number and different in kind ; 

 the conditions are in every way more re- 

 stricted and more absolute than elsewhere. 

 Hence we have here in an intense degree all 

 the known factors of species-making. On 

 the one hand, spontaneous variations are 

 more likely to occur through change of food 

 and circumstances ; on the other hand, selec- 

 tive action must be exerted in a very special 



