No. 105.] 217 



says, " although it has generally occasioned an increase of straw^ 

 the yield of grain has not been improved, and the crops have in ma- 

 ny instances been found unusually subject to mildew." Similar ex- 

 periments, by the writer, have produced no favorable result. Hence 

 we perceive that supplying, simply, an essential ingredient, does not 

 always answer the purpose. Artificial guano, made by an obser- 

 vance of the analysis of the natural, though useful, has not been 

 found nearly so powerful as the latter. Nitrogen, supplied properly 

 to plants, causes a healthy and rapid growth ; yet although this ele- 

 ment exists uncombined as a component of the atmosphere, and in 

 direct contact with the leaves of plants, they will perish for want of 

 it before they will draw a particle of it from the air. Hence in all 

 chemical deduction relative to manners, the experiments of the cul- 

 tivator only are to be depended on, and to remain as the decisive 

 test. Suggestions of incalculable importance may come from theory, 

 but practice alone must prove their value. 



The importance of the analysis of soils, to determine deficient in- 

 gredients, and then to supply defects, has been already adverted to. 

 Although its value thus appears to be very great, and has been much 

 extolled by chemical writers and their imitators, yet there are diffi- 

 culties in practice which render extreme caution in drawing conclu- 

 sions very necessary. The constituents of plants may indeed be 

 determined with much accuracy ; and the different ingredients in 

 manures, and their consequent adaptation to those plants, and of 

 their comparatively fertilizing effects, may be ascertained frequently 

 in the laboratory. But the extensive diffusion of these ingredients 

 through broad acres of soil, and the exceedingly minute proportion 

 which some bear to the whole bulk of the soil, renders the determi- 

 nation of these proportions, if not the actual existence of the ingre- 

 dients, difficult if not impossible. A distinguished chemist told the 

 writer, that for ordinary earthy substances, the detection of a thou- 

 sandth part required skilful analysis. Minuter portions of some 

 constituents are more easily detected than of others. But suppose a 

 ten-thousandth part the utmost limit for agricultural practice, a few 

 instances will show the inadequacy of analysis in cases which may 

 occur : 



A considerable portion of sulphate of lime or gypsum is found to 

 exist in red clover, and other leguminous plants. Hence a reason 



