396 CORRELATION OF DISTRIBUTIONAL FEATURES. 



between 0.40 and 0.50. There are, however, Grassland stations which 

 have readings as high as those of the minimum values for the Transi- 

 tion region. The maxima and minima of the Transition lie within 

 those of the Grassland and the Deciduous Forest, as is shown by the 

 simpler form of graph which we have generally used (figs. 21 to 74). 

 The maximum number of stations for both the Transition and the 

 Deciduous Forest is in the group having values of 1.00 to 1.10 for the 

 moisture ratio. There are only a few stations in the Deciduous Forest 

 for which the readings are higher than in the Transition region. It is 

 impossible to draw definite conclusions from the graph presented, but 

 it may be seen that this method of expressing results would be valuable 

 for cases in which data had been secured for evenly spaced localities. 



The climatic extremes given in the accompanying tables have been 

 derived from all of the climatic maps sufficiently detailed for this use. 

 In view of the large number of values involved it seemed desirable to 

 attempt a means of giving a graphic presentation of a portion of these 

 data. This has been done by constructing the graphs shown in figures 

 21 to 74. The manner in which these were made is such as to show 

 the relation between the extremes for each plant or vegetation and 

 the extremes for the United States as a whole. The graphs were made 

 by constructing a special scale for each one of the climatic charts. All 

 scales were of the same length and each bore at its left end the minimum 

 value. The scale was then marked so as to read in convenient units 

 the successive values that might be expected at different locahties in 

 the United States. To give a concrete example: The lowest and 

 highest values of the normal daily mean precipitation (plate 46) are 

 0.009 and 0.199 inch respectively. These values were placed at the 

 left and right ends, respectively, of the scale for this climatic feature. 

 The length of the scale in millimeters was then divided by the difference 

 between the extremes, 0.190 inch, and the length of scale was calcu- 

 lated that would express 0.010 inch. With this length as a unit the 

 scale was then subdivided by 19 lines, with calculated allowance for 

 the fact that the scale neither started nor ended with the even tens. 

 Out of the 31 sets of climatic extremes given in the tables, 17 were 

 selected as most important, and scales were made for each of them by 

 the above method. The graphs were then constructed by using the 

 appropriate scale to mark off the distances in each block that would 

 express the extremes shown in the table. 



Two kinds of graphs have been made, the first of which (figs. 21 to 

 26) show the limits of the same climatic feature for each of the general- 

 ized vegetational areas of the United States. In these graphs each 

 block has been marked off by the same scale, and the maximum and 

 minimum values are therefore given at the top of the graph. In the 

 remaining graphs (figs. 27 to 74) are shown the extreme values of 



