CORRELATION OF DISTRIBUTIONAL FEATURES. 573 



Deciduous Forest occupies the remainder of the eastern subdivision of 

 the semihumid province, so that if the two last-named vegetation tj^pes 

 are grouped together they occupy the whole eastern subdivision of the 

 semihumid province. Northeastern Evergreen Forest occupies very 

 nearly the northern portion of the eastern subdivision of the humid 

 province, but this vegetation type extends southward in the Appa- 

 lachians, which is not shown for the corresponding climatic area by the 

 precipitation-evaporation ratio; but this southward extension of the 

 northeastern humid conditions is shown by the line for 140 on the 

 evaporation chart (plate 53, figs. 3 and 22). Southeastern Evergreen 

 Forest occupies nearly the same area as the southern portion of the 

 eastern subdivision of the humid province. 



This agreement is about as close as could be hoped for in work of this 

 kind, and the correlations are very nearly what might have been 

 expected. Two apparently important features require brief mention: 

 (a) the relation of the Deciduous Forest area to that of the Deciduous- 

 Forest Grassland Transition, and (6) the relation between the climatic 

 conditions concomitant with the Northeastern and Southeastern Ever- 

 green Forests. 



(a) It will be noticed that no one of the climatic maps shows any 

 line that may be considered as approximating the position of the 

 boundary between the deciduous forest and the prairie. "WTiile this 

 boundary, like the other lines of plate 2, does not represent a sharp 

 line of demarcation, nevertheless it is one of the most pronounced 

 and clearly recognizable vegetational boundaries presented by the 

 United States. It is actually a simple matter — for example, in 

 Minnesota, Indiana, or Illinois— for an observer to step, within a 

 very few meters, from what is clearly and unequivocally decid- 

 uous forest into what is just as unquestionably prairie. This is 

 not nearly so easy in the case of the other vegetation boundaries as 

 actually encountered in the field; deciduous and evergreen forest 

 usually mingled near their common margins, and desert, grassland, 

 and prairie usually intergrade quite imperceptibly, so that their bound- 

 aries frequently have to be regarded as bands or zones many kilo- 

 meters wide, even by an observer in the field. Furthermore, various 

 species of trees have recently been introduced upon the upland of the 

 prairie region, which originally was forested only on the flood-plains 

 of the streams, while the deciduous forest of Pennsylvania, Ohio, etc., 

 has been largely removed, so that the general aspect of the country is 

 now much the same as in Iowa or eastern Kansas and Nebraska. This 

 fact has led many students to regard the prairie region as potentially a 

 deciduous-forest region, as far as climatic conditions go, and various 

 non-climatic conditions have been suggested as explaining the original 

 absence of trees from the prairie uplands. 



Looked at from the dynamic standpoint, it seems clear to us that the 

 difference in environmental conditions that has to be postulated as 



