CORRELATION OF DISTRIBUTIONAL FEATURES. 519 



plants more precisely than the ratio of precipitation to evaporation. 

 In dealing with large areas, however, the soil-moisture is so closely a 

 function of the precipitation that the two expressions frequently 

 approach identity, or at least proportionality. Of course, it is under- 

 stood that in all cases where the soil-moisture content (percentage of soil- 

 moisture on a volume or weight basis) is employed as the index of the 

 soil-moisture condition, great differences in the water-retaining power 

 of the soil in different areas must upset this clear relation. If the soils 

 compared are all clay or all sands, etc., the soil-moisture content itself 

 is probably a relatively good index of the water-supplying power of the 

 soil, but this is not true when sands are to be compared with clays, for a 

 10-per cent water-content in a sand may be physiologically equivalent 

 to a 50-per cent content in a clay, other conditions being considered as 

 alike. 



IV. CONDITIONS THAT PROBABLY DETERMINE THE LIFE-ZONES OF 



MERRIAM. 



I. OBSERVATIONS FROM THE CHARTS. 



In a discussion of the climatic extremes of the life-zones of the 

 United States, as outlined by Merriam, it is necessary for us to dis- 

 regard the Boreal and Tropical Regions, and also the Gulf Strip of the 

 Lower Austral Zone, on account of the very small number of climato- 

 logical stations comprised in those areas. As originally drawn by 

 Merriam, the life-zone map of the United States was in actuality a 

 climatological map, based on a summation of temperatures and 

 slightly modified, particularly on the Pacific coast, by data on the 

 temperature of the hottest 6 weeks of the sunamer. It is not necessary, 

 therefore, to make correlations of the range of the life-zones with the 

 distribution of the different intensities of the remainder summation of 

 temperatures above 32° for the year (plate 37, which is Merriam's 

 chart), nor with the normal daily mean of the hottest 6 weeks of the 

 year (plate 44, which is also Merriam's). 



In figures 37 to 42 are shown the graphs for the leading climatic 

 dimensions given in tables 33 to 38 for the Transition, Alleghanian, 

 Upper Sonoran, Carolinian, Lower Sonoran, and Austroriparian zones. 

 These 6 subdivisions comprise the western or arid and the eastern or 

 humid subdivisions of the transcontinental zones based on temperature 

 conditions. They represent, in other words, an exact but special sub- 

 division of the country on a basis of certain temperature conditions, 

 together with a rough subdivision on the basis of moisture conditions. 

 The line separating the arid and humid divisions of these zones is 

 drawn by Merriam along the one-hundredth meridian from Oklahoma 

 to South Dakota, departing a Httle to the eastward at the north and 

 south. This line corresponds rather closely with the isoclimatic line 

 of a value of 0.60 for the moisture ratio tt/E. It does not take account. 



