466 Ike American Naturalist. [June, 



controlling. To understand the vegetation of a region one 

 must ascertain not only what are its physical, meteorological 

 and geological features, but much more what sorts of plants 

 control its water, meadow, plain, or forest vegetation. Directed 

 towards the latter ends, statistics have a very different mean- 

 ing. Such work is the aim of the new geographical botany. 

 " When we hear of a district," say Schroeterand Stebler, " that 

 it is covered with extended fields of turf-rush or of brome- 

 grass, that tells us more of the nature of the region than long 

 lists of meteorological data. It also tells us more than the 

 mere occurrence of the species in question of itself " 



A notable contribution to this department of the science is 

 Dr. Drude's new work, " Deutschlands Pfianzengeographie," 

 of which the first part appeared in January last. The sub 

 title of the work gives a clue to its purpose. It is stated to 

 be "ein geographisches Charakterbild der Flora von Deuts- 

 chland." Much has been done in recent years towards such 

 characterization of restricted districts, or for large areas as 

 regards certain kinds of vegetation. But Dr. Drude in giving 

 a complete picture of the vegetation of as large a country as 

 Germany has, in one sense, made an epoch in geographical 

 botany. Such a work demonstrates that the era of preparation 

 is passed. A mere cursory examination of the work serves to 

 convince the reader that the theory and system of plant- 

 geography have been thoroughly worked out, and that hence- 

 forth workers will be busied chiefly with their application to 

 other regions rather than with devising new methods. 



As has been remarked, in order to be of value, statistics must 

 be based not upon the systematic groups of plants but upon 

 groups founded on biological considerations, so far as they 

 indicate a positive role in the vegetation of the region in ques- 

 tion. Such groups are called vegetation-groups. Dr. Drude 

 points out also that the proportions of the number of repre- 

 sentatives of the several orders, genera, or other systematic 

 groups are not to be reckoned with the whole flora of a region 

 as represented by a certain number of species, but with the 

 biological plant-community of the region. Accordingly he 

 constructs some thirty-five vegetation-groups for the flora of 



