No. 468.] AMERICAN FOREST CENTERS. 883 
Rhus glabra, R. aromatica, R. radicans, Cornus asperifolia, C. 
amomum, and Xanthoxylum americanum. This society is of 
very different origin and represents the western border of the 
shrub stage of the Deciduous forest formation. The shrubs are 
succeeded by trees belonging to the same center, among which 
are Juniperus virginiana, Ostrya virginica, Celtis occidentalis, 
Quercus acuminata, Cercis canadensis, Ulmus fulva, etc. A 
very different combination of societies would be met with in 
similar situations in northwestern Minnesota or southeastern 
Texas because of the different centers involved. 
The mapping of these centers naturally brings up the question 
of the climatic determinants of each. During glacial times the 
Northeastern Conifer must have been mixed with the Decidu- 
ous forest. Why are they so distinctly separated at the present 
time? What are the causes of the * prairie peninsula " in Iowa, 
Illinois, and Indiana; and the region of open forests adjoining it ? 
Naturally we look for some method of mapping climatic data, 
which will show climatic centers in approximately the same 
positions as the centers of plant distribution. An examination 
of monthly, seasonal, and annual temperature and rainfall maps 
shows that neither of these factors alone can do this. Historical 
considerations may aid in explaining the relative positions of these 
centers, but are inadequate for the complete explanation of their 
present limits. 
A method was accordingly sought by which temperature and 
moisture data could be combined in a single number. The fact 
that so large a part of all plant adaptations is directly or indi- 
rectly connected with transpiration, suggested that if the ratio 
of the rainfall to the evaporation were determined, a new basis 
for mapping would be at hand which would involve several cli- 
matic factors. The depth of evaporation depends upon (T. Rus- 
sell, '88) the temperature of the evaporating surface, the relative 
humidity of the air, and the velocity of the wind. These are the 
same climatic factors which most powerfully affect transpiration, 
and which must be of great importance in determining the geo- 
graphic range of plants. Unfortunately, the only figures for evap- 
oration available are those published by T. Russell for the year 
beginning July 1, 1887, and extending to July ı, 1888. They 
