STATISTICS OF METASPERMAE. 



643 



necessary. Preponderance of species in one of the regions 

 over the opposite region is taken as the index of range. 

 The 95th meridian W. of Greenwich is adopted as the line di- 

 viding the eastern from the western half of the continent, and 

 the 45th parallel of N. latitude as the line dividing the north- 

 ern from the southern region. If then in a given genus a greater 

 number of species occur north and east of the dividing lines 

 than south and west, such a genus is entered as northern and 

 eastern, in the table. No attempt at hairsplitting distinctions 

 has been made, and genera developed pretty evenly in all parts 

 of the continent are entered in each of the four groups. Very 

 commonly a genus will be entered in three groups. The groups 

 of three which are most common are the north- east- west, the 

 south-east-west and north-south-east groups. It is believed 

 that by such a comparatively elastic method of entry more ac- 

 curate results will be obtained than if one were to attempt for 

 each genus to strike such a demarcation line that it could fall 

 into but two groups. Again, in the case of genera either 

 monotypic or with very few species, the numerical test must be 

 abandoned. t In its place is adopted the specific range and the 

 comparative frequency of individuals on different sides of the 

 demarcation lines. From these considerations it will be seen 

 that the north-south-east-west group of genera is of the most 

 generally distributed genera while the south-east or north-east 

 group and others of similar construction are the least gener- 

 ally developed in the North American continent. Large de- 

 velopment of range may however, and often does, exist to- 

 gether with small development of species, or vice versa. With 

 so much of emphasis upon the elasticity of grouping the table 

 may be presented. 



C. Table Illustrating Continental Development of Genera 

 Represented in the Minnesota Talley. 



