180 On the Geographical DL^trWutiou of Insects. 



These figures, which express the numerical relations of 

 groups to each other, by no means afford an exact idea of theii* 

 real distribution. A map alone could do this, if it were pos- 

 sible to construct one. In fact, every group (with a few excep- 

 tions), whatever may be its importance, has a certain point of 

 the globe where it becomes dominant, that is to say, where 

 its elements are assembled in greater numbers than in any 

 other place ; on leaving this centre, it sends out rays or branches 

 in different directions, which are the more numerous, and 

 extend in general to greater distances, the higher in value 

 the group happens to be ; in other words, the ramifications of 

 a family are more prolonged than those of a tribe, those of a 

 tribe than those of a genus, and so on. These branches are 

 immediately connected with the trunk when the area is conti- 

 nuous, or they are separated from it, or even occasionally 

 broken into numerous parts, when it is di.sjoined. Each gi'oup, 

 besides, has what may be called its strayed children, which 

 ai'e found widely insulated from their associates, and wander- 

 ing among other groups with which they have no affinity in 



