X.] PURPOSE OF SOIL SURVEYS 293 



much in evidence that the "typical soil" can only be 

 constructed by putting together the results of many 

 separate determinations. 



But the practice of constructing a typical soil for 

 analysis by mixing together equal fractions of many 

 samples drawn in the area in question, is not to 

 be recommended. Not only may an entirely foreign 

 or accidentally impure sample be introduced without 

 detection, but, further, the limits of variation normally 

 to be expected in individual soils of the same type is 

 just as important as the composition of the type itself. 

 Again, the existence of unsuspected systematic varia- 

 tions is entirely obscured by any process of mixing 

 samples. The character of the information which 

 should accompany the soil maps must largely depend 

 on the purpose of the survey, whether it is concerned 

 with the agriculture of an old and settled country, or 

 whether it partakes of the nature of an exploration, and 

 aims at showing the capacities of the land for new 

 crops and industries. In the United States, for 

 example, the latter form of soil survey is exemplified ; 

 in many parts of the country agriculture is so recent 

 that there is no accumulation of experience as to the 

 crops most suited to each kind of land; hence the 

 survey, by comparisons of the texture of the soil, the 

 climatic conditions, and the depth to ground water, 

 with the conditions prevailing in better known areas, 

 can directly tell the settler with what crops he is most 

 likely to succeed. The cultivation of special crops like 

 tobacco and sugar beet, to take two examples of special 

 interest at the present time in the United States, 

 can be extended into new districts possessing suitable 

 soils, with a minimum of the risk which must always 

 attend the introduction of a novel form of culture. The 

 suitability of other classes of land for irrigation, the 



