376 SOILS, 



meter size also creates an unfortunate and ineliminable dis- 

 crepancy. 



In order to exhibit clearly the influence of climate as distinct 

 from other local conditions, it was also necessary to eliminate, 

 in both the arid and humid regions, the soils directly derived 

 from, or connected with calcareous formations; such as the 

 prairies of the Southwestern States, the Bluegrass region of 

 Kentucky, etc. This rule having been applied impartially to 

 the soils of both climatic regions, it can hardly be questioned 

 that the conclusions flowing from a discussion of the results of 

 the comparison are entitled to as much weight as are those of 

 any comparison based on large numbers of observations made, 

 not with reference to the special point under consideration, but 

 with a practical object of which the governing conditions were 

 more or less uncertain, and required to be ascertained by a pro- 

 cess of elimination. 



The table gives, first, the averages for each ingredient for each of the 

 states represented, the number of analyses from which the averages are 

 derived being given in each case. These averages are given separately 

 for the states of the humid and the arid regions respectively ; and at 

 the base of each group the grand average is shown in two forms. The 

 first gives the figures as derived from the aggregate number of soil 

 analyses in each great group, being 696 for the humid, 178 for the 

 transition region and 573 for the arid, divided into the totals resulting 

 from the summation of each ingredient for the whole 696, 178 and 573, 

 respectively. 



The second form is that in which the soils of each state are considered 

 as representative of the general character of such state, as the result of 

 intentional selection ; such as actually occurred in the cases of those 

 included in the census work of 1880. The figures given here are 

 therefore the result of a summation of the state averages as such, and 

 of their division by the number of states represented. 



It will be noted that while these two modes of presentation do change 

 the figures a little, yet in either form the same general result is out- 

 lined with striking accuracy. It is also notable that notwithstanding 

 the less complete extraction of soil-ingredients in the starred states, 

 the general ratios between arid and humid soils remain substantially 

 the same. For Western Oregon, local calcareous formations compel 

 omission of three lime figures from the averages. 



