SOIL ANALYSIS AND ITS INTERPRETATION 323 



basis of the whole work is empirical : the agricultural and 

 vegetation characteristics have first to be ascertained by field 

 trials, and then systematised and amplified by aid of the 

 laboratory data. It is necessary to begin, therefore, by going 

 over the whole region very carefully and dividing it up in 

 areas within which similar agricultural or vegetation character- 

 istics prevail. In general the areas differentiated in the 

 geological drift maps will be found identical with the vegeta- 

 tion areas, especially if the drift map is interpreted in the 

 light of the Memoirs of the Geological Survey. But as agri- 

 culture is influenced by altitude and rainfall, the investigator 

 must also use an ordinary contour map and a rainfall map, 

 which, unfortunately, he must construct for himself from data 

 in British Rainfall. Within each vegetation area a number 

 of soil samples must then be taken from points representing 

 the area as closely as possible : situated, for example, on level 

 regions or long gentle slopes, and not on made ground, steep 

 slopes, or places of local disturbance, etc. 



In seeking for typical places the investigator is likely to 

 meet with a good deal of discouragement ; farmers will tell 

 him that half a dozen or more different kinds of soil occur on 

 their particular farms, while the vegetation of a wild area may 

 show considerable changes. But these differences often arise 

 from differences in the compound particles and not in the 

 ultimate particles ; small variations in the amount of calcium 

 carbonate or of organic matter, or differences in the water 

 supply, or management, may considerably affect the ease of 

 cultivation and the vegetation relationships and give the im- 

 pression of a wholly different type of soil. So little does 

 cropping, cultivation, etc., affect the ultimate particles that it 

 is quite immaterial for the purpose of a soil survey whether the 

 sample is taken from pasture land or arable land, but it is well 

 to take a number of samples from both. With a little practice 

 abnormal places are easily avoided. It is necessary to take a 

 large number of samples, not only to get at the type, but also 

 to trace out the causes of the phenomena noted by the farmer, 

 and to discover the main factors determining the cultivation 



