8 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 65 



type, after its first appearance, whether by original generation or 

 introduction from outside, will (as in part suggested by the be- 

 havior of introduced species) be marked first by individual variation, 

 soon leading to more or less fixed varieties, and finally to the evolution 

 of new species and even new genera, each of which was originally the 

 exponent of conditions more or less different from those under which 

 the type originally appeared. 



Now in most large genera we find among the component species one 

 which in its characters occupies the mean between the extremes shown 

 by the other forms, and which typically covers the entire economic, 

 physical, and geographical range of the genus, unless the species on 

 the borders of the generic range are isolated by barriers. 



Obviously this is the species best adapted to the conditions of the 

 present day and, if conditions should remain indefinitely as they are 

 now, such a species would gradually succeed, by the mere force of 

 numbers and greater procreative power, which have already enabled 

 it to overrun all the other forms, in exterminating all of the other 

 species of the genus which it was able to reach. 



As families and orders are constructively the same as genera, we 

 typically find in them a highly dominant genus, subfamily, or family, 

 which stands in the same relation to them that the dominant species 

 does to the genus. 



And among the higher groups the same thing is repeated ; thus, for 

 example, we find among the mammals the rodents, among the birds 

 the finches, among the fishes the perches, among the flies the muscoid 

 types, etc., each group including species almost all of which are of 

 small average size, yet never excessively small, representing the 

 dominant types which appear to be on the road to supplanting all the 

 other types through a development from their immediate stock of 

 virile competing forms, and which, were conditions to remain in- 

 definitely as they are at the present epoch, would eventually come to 

 form the entire world fauna. 



An appreciation of the normal existence of such a dominant type 

 in each large and widely distributed group is essential for the compre- 

 hension of the fact that, given a number of closely related genera 

 occupying a large area, but separated from each other by barriers, the 

 genus occupying the center of distribution will be the most special- 

 ized, while that at the periphery will be the least specialized. 



Let us suppose a genus recently arisen, occurring uniformly over 

 a very wide area in the center of which the conditions grade very 

 slowly from the optimum to impassible physico-economic barriers in 

 each direction, while at the periphery the conditions grade very 



