THE PHYSICO-CHEMICAL INVESTIGATION OF SOILS. 



325 



are partly or wholly due to other causes than those purposely 

 supplied by the experimenter. 



Analysis of Cultivated Soils, It is also clear that in view erf the in- 

 evitable complexity of the conditions governing vegetable growth, we 

 should whenever feasible proceed from the more simple to the more 

 complex. The failure to conform to this rule in soil investigation has 

 been the cause of an enormous waste of energy and work bestowed, at 

 the very outset, upon the most complex problem of all, viz., the investi- 

 gation of soils long cultivated and manured ; lands which, having been 

 subject perhaps for centuries to a great and wholly indefinite variety of 

 crops and cultural practices, had thereby become so beset with artificial 

 conditions that without a previous knowledge of what constitutes the 

 normal regime in natural soils, the correlation of their chemical consti- 

 tion, as ascertainable by our present methods, with their production 

 under culture, became as complex a problem as that of motions of three 

 mutually gravitating points in space. Neither can be solved by the 

 ordinary processes of analysis, chemical or mathematical. Nevertheless, 

 though it was at one time contended that the minute proportion of 

 plant-food ingredients withdrawn from soils by cultivation could not be 

 detected by quantitative analysis, numerous examples have shown that 

 with our present more delicate methods this can in most cases be done, 

 though not always after a single year's crop. 



Mftliods of Soil Analysis. The more or less incisive solvent agents 

 used in extracting a soil for analysis will of course produce results 

 widely at variance with each other. When fusion with carbonate of 

 soda, or treatment with fluohydric acid is resorted to, we obtain for 

 each soil-ingredient the sum of all the amounts contained in each of 

 the three categories the unchanged minerals, the zeolitic " reserve," 

 and the water-soluble portion. It was early recognized that the results 

 of such analyses bear no intelligible relation to the productive capacity 

 of soils; for pulverized rocks of many kinds, or volcanic ashes freshly 

 ejected and notoriously incapable of supporting plant growth, might be 

 made to give exactly the same composition. The amounts of plant- 

 food ingredients thus shown might be several hundreds or thousands of 

 times greater than what one crop would take from the soil, and yet not 

 an ear of grain could be produced on the material. The only case in 

 which any useful information could be thus obtained would be that of 

 the absence, or great scarcity, of one or more of the plant-food in- 

 gredients. 



