574 CORRELATION OF DISTRIBUTIONAL FEATURES. 



related to the difference in vegetation here considered must be rather 

 recondite and subtile in its nature. We are, however, strongly inclined 

 to maintain that this environmental difference will prove to be a cli- 

 matic one, though probably not measurable in terms of any of the 

 simpler climatic indices. Here is a problem that is well worthy of 

 much deeper study than we have been able to give it. We wish to 

 suggest one possible dynamic explanation on climatic grounds. 



If our charts showing mean daily evaporation (plate 53, figs. 3 and 

 22) be once more examined, it will be recalled that the eastern sub- 

 division of the semiarid moisture province here exhibits a great eastern 

 lobe reaching from Oklahoma to Pennsylvania. This general phe- 

 nomenon is shown or suggested on other moisture charts, and may be 

 tentatively regarded as of climatic significance, until more thorough- 

 going studies of the aerial moisture conditions become possible. Now, 

 this penetration of semiarid conditions into the center of the great 

 eastern area of the semihumid province suggests that the explanation 

 of the vegetational transition before us is probably largely related to 

 evaporation. The same conclusion is suggested by the relative air- 

 humidity chart (plate 65, figs. 17 and 24). Just how atmometric or 

 air-humidity data should be treated in order to obtain a moisture 

 index that may bring this point out in a satisfactory way, if it be true, 

 can not, of course, be predicted. In support of the general probability 

 that evaporation is the main climatic feature to be called upon to 

 explain this vegetational transition, it should be remarked that small 

 local prairies were of frequent occurrence in Indiana and Ohio when 

 these regions were still under forest, so that the tension zone between 

 forest and prairie was apparently very broad in the region south of the 

 Great Lakes. It should also be mentioned that the evaporation data 

 used for these studies all refer to a single year (Russell's data, 1887-88) 

 and it is suggested that a normal evaporation chart may show an iso- 

 atmic line approximating the position of the prairie-forest boundary 

 here in question. 



(6) The Northeastern and Southeastern Evergreen Forest types of 

 vegetation are well known to be very distinct, at least floristically; yet 

 they are to be regarded (along with the Western and Northwestern 

 Evergreen Forests) as physiologically or ecologically rather similar, 

 being dominated by evergreen, needle-leaved trees. It is therefore 

 important to note that the climatic conditions that seem to correspond 

 to the Northeastern Evergreen Forest are no more continuous with 

 those corresponding to the Southeastern Forest than is the actual dis- 

 tribution area of the former forest itself with that of the latter. This 

 point is clearly shown on the three charts for E, P/E, and H (plates 

 53, 57, and 65; figs. 2, 3, and 17, and 21, 22, and 24), and a still more 



