GEOGR.'VPHIC ISOLATION AND SPECIES-FORMING 119 



but not derived from one anotlier by direct lines of descent. A 

 closer study indicates that some of them '^camc from closely 

 related forms in remote geographic areas, others from antece- 

 dent forms now extinct, and not more than three or four from 

 species still inha])iting the region/' 



Tho nature of any fauna bears an immediate relation to 

 the barriers, geographic, climatic, topograi)hic, or bionomic, 

 which may form its boundaries. By bionomic barriers wo 

 mean any condition of an}^ sort which may check free inter- 

 bi-eeding, or which may tend to cause divergence within a 

 species. A thickly peopled level area may be in this sense a 

 barrier, because it prevents the animals on the one side of the 

 area from interbreeding with those on the opposite side. If 

 the two extremes have diverged to become different species, 

 the individuals in the middle area, whose presence in a sense 

 constitutes the bionomic barrier, are usually variously inter- 

 mediate in the characters and habits which they possess. 



Whenever the individuals of a species move evenly over an 

 area, its members freely interbreeding, the character of the 

 species remains substantially uniform. Whenever freedom of 

 movement and consequent freedom of interbreeding is checked, 

 the character of the species is rapidly altered. It is changed 

 even though external conditions seem to be absolutely identical 

 on both sides of the barrier, and if there is no visible distinction 

 in the original stock on the two sides. Presumably, there are 

 subtle differences in the environment, producing changes in 

 the process of selection and adaptation. Doubtless, there are 

 differences ec^ually subtle produced by the processes of varia- 

 tion and their repetition by inheritance. 



The pregnant phrase of Dr. Coues applies in these cases: 

 "Migration holds species true: localization lets them slip." 

 In other words, free interbreeding swamps incipient lines of 

 variation, and this in almost every case. On the other hand, a 

 barrier or check of any sort brings a certain group of individuals 

 together. These are subjected to a selection different from 

 that which obtains with the species at large, and under these 

 conditions new forms are developed. This takes place rapidly 

 when the conditions of life are greatly changed, so that a new 

 set of demands is made on the species, and those individuals not 

 meeting it are at once destroyed. The ])r()cess is a slow one, 

 for the most part, when the barrier in question interrupts the 



