Mineral Plant Foods contained in Soils. ' 351 



instances obscure. Again, a large number of the analyses are 

 of samples sent in by farmers to the various State experiment 

 stations in the hope that a chemical analysis would shed a 

 light on some peculiarity of a, perhaps, isolated and. untypical, 

 small area of soil. In either case such samples must be con- 

 sidered as being individually unrepresentative of the region of 

 their origin. 



The analyses given by Merrill* show that, in humid regions, 

 rock weathering results in a greater proportional loss of the 

 relatively soluble constituents of rock than of the less soluble 

 substances, that is, primarily of silica. In other words, residual 

 soil contains a lesser percentage of calcium and phosphorus 

 than the rock material from which it was derived. An 

 extreme case is that of limestone residual soils which are often 

 almost totally deficient in lime. 



The facts exhibited by Merrill's analyses of specific rock types 

 and the residual material derived from them can be more 

 generally shown by a consideration of the averaged bulk 

 analyses of representative series of soils, contrasted with the 

 average chemical composition of the rock material composing 

 the lithosphere. Such analyses are available in publications 

 by Clarke,-]- Veitch,;}: and Mooers.§ Clarke has calculated the 

 average composition of the outer ten-mile shell of the litho- 

 sphere. Veitch presents a comprehensive series of bulk 

 analyses of Maryland soils, comprising residuals from many 

 rock types, both igneous and sedimentary, and of greatly 

 varying geological age. The samples he analyzed were from 

 the subsoil immediately below the top soil ; and each analysis 

 represents the average of several determinations from different 

 localities. Considering the wide range of the original rock 

 formations represented, the averaged values of the analyses, 

 and the nature of the material selected for analysis, it is fair to 

 assume that an average of these averages will, in an indicative 

 way, express the mean composition of the soil material residual 

 from weathering and erosion processes in a temperate, humid 

 region. Mooers' results may be considered as confirmative of 

 the averaged figures from Veitch. They are less representative 

 than the Maryland series, especially in that eight of the ten 

 soils analyzed, and here averaged, were derived from lime- 

 stones and dolomites. Below are tabulated, in percentages, 

 the averaged amounts present of three important mineral 

 plant food constituents, as obtained from these investigations : 



* Merrill, G. P., Eoeks, Rock Weathering, and Soils, Macmillan Co., 1906. 



t Clarke, F.W., Data of Geochemistry, Bull. 330, U. S. Geol. Survey, 1908, 

 p. 31. 



X Veitch, F. P., The Chemical Composition of Maryland Soils, Bull. 70, 

 The Md. Agri. Expt. Station, 1901. 



§ Mooers, C. A., The Soils of Tennessee, Bull. 78, Agri. Expt. Station of 

 the University of Tenn., 1906, p. 74. 



