§ 108. EXPERIMENTS WITH PLANTS AND SOILS. 203 



the quantity or quality of the crop, or whether, as is 

 probably the case, the proportion between the clay and 

 these soluble substances, or, in other words, the physical 

 character of the soil, exerts a controlling influence on its 

 fertility. 



It would be well to perform three experiments of the 

 same character with each kind of soil, partly for the pur- 

 pose of securing greater certainty in the results, and 

 partly that the same soil may be used afterwards for other 

 experiments. 



To test the question whether the substances in the soil 

 tliat are soluble in pure or carbonated Avater exert any 

 essential influence on its immediate productiveness, a trial 

 may be made with a soil in its natural condition, and with 

 the same or a similar soil, after it has been exhausted with 

 water in the manner described for the preparation of an 

 aqueous solution for chemical analysis (§ 100). As, how- 

 ever, it would be very inconvenient to exhaust such a 

 quantity of soil in this way, as would be required to fill 

 one of the boxes, this experiment may be performed with 

 but 7-10 kilos, of soil, in smaller boxes, or in glass 

 vessels. The plants should be watered with distilled 

 water during the progress of the experiment. 



Valuable results may be obtained from experiments in 

 M'hich equal quantities of assimilable plant-food are added 

 to diflerent varieties of soil. 



The action of a full and complete provision of the ele- 

 ments of plant-food on one kind of plant grown in the 

 diflerent soils, should first be examined. For such a 

 complete manuring, we need only to mix together acid 

 potassic phosphate, calcic nitrate, potassic nitrate, and 

 magnesic sulphate, so that the relative proportions of 

 the bases in the mixture will be the same as in the ash 

 of the plant to be cultivated. 



For this purpose, the boxes and soils employed in pre- 

 vious years for trials with soils containing different 



