62 FORREST SHREVE 



as in the deserts of Arizona, and the dehmitation of plant com- 

 munities with a reasonably constant composition is even more 

 difficult than in the desert. In the forest and grassland areas of 

 temperate climates the natural communities of plants are more 

 simple in their composition and more distinctive in comparison 

 with adjacent communities. From the standpoint of succes- 

 sional ecology every plant community, wherever found, con- 

 stitutes a unit, — a definite entity. From the standpoint which 

 I am desirous of taking, each of these communities is a collec- 

 tion of species with some very general resemblances and some 

 very pronounced dissimilarities in their physiological character- 

 istics. The findings of successional ecology will not aid us in 

 learning the nature or intensity of the controls which limit a 

 species to a particular geographical range or to a particular class 

 of habitats. 



In order to face squarely the problems of causation in both 

 general and local plant distribution it is necessary to consider 

 plants as physiological entities and not as phylogenetic entities. 

 It is also necessary to give our most serious consideration to 

 individual plant species rather than to plant communities. Our 

 understanding of the limitation of general distribution, and of 

 habital distribution as well, is dependent upon a fuller knowledge 

 of the behavior of each of the physiological entities separately 

 considered. Our physiological knowledge, which is so strong 

 with respect to wheat and corn, Vicia faba and Lupinus albus, 

 must be extended to a greater familiarity with the dominant 

 plants of our leading vegetations, to a comparative knowledge 

 of the behavior of plants which grow together in the same com- 

 munities, and to a determination of the limiting factors which 

 control the individual or collective physiological functions of 

 the plant. 



There are two important phases in the present physiological 

 trend of work in plant geography and ecology. One is the estab- 

 lishment of correlations between the distribution of physical 

 factors and the distribution, general or local, of plant species 

 and communities. The other is the intensive study of a par- 

 ticular function, or group of functions, in a series of species which 



