398 CORRELATION OF DISTRIBUTIONAL FEATURES. 



In a few cases the comparison of graphs has been facilitated by using 

 double blocks. This has been done to show the comparative condi- 

 tions of the Deciduous Forest region and the Evergreen Needle- 

 leaved areas considered collectively (fig. 36), and to show the compar- 

 ison between the climatic extremes of plants of very different range, 

 in the cases of Sapindus marginatus and Populus halsamifera, and of 

 Cornus canadensis and Spermolepis echinatus. 



It is difficult to devise any means of graphically representing the 

 conditions that accompany the zones of diminishing abundance in 

 the case of the charts for the cumulative occurrence of plants of the 

 same growth-form. Three diagrams have been prepared to show the 

 amplitudes of conditions in these cases. One is for the 13 species of 

 commonest eastern deciduous trees, showing the conditions for the 

 areas that have 8 to 12 and 1 to 7 species, respectively. This graph 

 has triple blocks, each of which was constructed by use of the scales 

 that have been described, and the significance of the shading is ex- 

 plained by the key. The other diagrams are for the zones of abund- 

 ance of Bulbilis and Liriodendron, and are also explained by keys. 



All of the graphs showing the range of climatic conditions would 

 have a much greater value if they could be drawn in such a way as to 

 show what part of the total range of a climatic index is characteristic 

 of the largest, best-developed, or most typical part of a given vegeta- 

 tional or distributional area. Several test graphs were prepared show- 

 ing by vertical lines in each block the values for each of the stations 

 in the distributional area selected. These graphs, however, proved 

 to be not so much an exhibit of the conditions prevailing through most 

 of the area as they were of the irregularity in the location of available 

 meteorological stations, a matter over which we of course had no 

 control. For this reason it has seemed unwise to present these graphs, 

 or to enter into a discussion of this aspect of the correlation problem. 



II. CLIMATIC EXTREMES FOR EACH OF THE VARIOUS 

 VEGETATIONAL FEATURES. 



In the following tables are presented the climatic extremes for all 

 of the botanical areas investigated, as obtained by the methods just 

 described. These cover the general vegetational areas, the life-zones, 

 the areas of relative abundance of growth-forms, the areas of relative 

 abundance of individual species (their ecological distribution), and the 

 distributional areas of individual species. In each table are given the 

 highest and lowest values for each of about 31 climatic conditions for 

 the botanical area in question. The numbers given in the first column 

 of each table refer to the plates on which the climatic data are given 

 in map form. The same numbers are also used in the figures shown 

 on subsequent pages (figs. 21 to 74). Temperatures are given in 



