39 On (he Analysis of Soils, 



analyst oucbt to be well grounded in chenVical niformation;' 

 but perhaps there is no better mode ot" gaining it, than that^ 

 of attempting orisiinal investigations. In pursuing his ex- 

 periments, lie will be continually obliged to learn from 

 books the history of the subsiances he i.* employing or act- 

 ing upon ; and his theoretical ideas wi-ll be more valuable in 

 being connected with practical operation, and acquired for 

 the purpose of discovery. 



XVr. On- the Imprcjveyneiit of Soils, as connecUd with the 

 Principle of' their CoTUposilion. 



In cases v, hen a barren soil is exaniined with a view to 

 its improvementj it ought in all eases, if possible, to be 

 compared with an extremely fertile soil in the same neigh- 

 bourhood, and in a similar situation : the difierence given 

 by their analyses would indicate the methods of cultivation j 

 and thus the plan of improvement would be founded upon 

 accurate scientiiic principles. 



If the fertile sail contained a large quantity of sand, iit 

 proportion to the barren soil, the process of amelioration 

 would depend simply upon a supply of this substance; and 

 the method would be equally simple with regard to soils de- 

 ficient in clay or calcareous matter. 



In the application of clay, sand, loam, marie, or chalk, to 

 lands, there are no particular chemical principles to be ob- 

 served ; but when quicklime is used, great care nmst be 

 taken that it is not obtained from the magnesian limestone; 

 for in this case, as has been shown by Mr. Tennant, it is 

 exceedingly injurious to land*; The magnesian limestone 

 may be distinguished from the conmion limestone by its 

 greater hardness, and by the length of time that it requires 

 for its solution in acids, and it may be analysed by the pro- 

 cess for carbonate of lime and magnesia IX. 



When the analytical comparison indicates an excess of 

 vegetable matter, as the cause of sterility, it may be de- 

 stroyed by much pulverization and exposure to air, by 

 paring and burning, or the agency of lately made quick- 

 lime. And the defect of aniryal and vegetable matter musfc 

 be surplied by animal or vegetable manure. 



XVII. Sterile Soils in different Climates and Situations 

 must differ in Composition. 



The general indications of fertility and barrenness, as 

 found by chemical experiments, must necessarily differ m 



* Phil. Transactions for 1799, p. !^0j. This limestone is found abundantly 

 in Yorksjiire, Derbyshircr and Someractshire. 



8 different 



