131 



The experiments condueled by Fraps, by Upson nnd 

 Powell, and by Davidson gave results which are in 

 ahnosl e()nii)lele agreement with our work; while the 

 tests by Schreiner and Skinner do not accord closely. 



Where soil culture exjxrinunts Tail to give concord- 

 iinl results, in all pi'ol)ability, sueh lack of agreement 

 is due to the diilerenees betwei-n tlie soil types used 

 ill the experiments. 



11 would seem clear tlial the soil toxin theory cannot 

 apply to the soils used in these experiments. In each 

 case, infertility of the check pots can be completely ex- 

 {)lained as being due to a lack of mineral nutrients. 

 1 1 is also evident from the data here })resented that 

 solution culture and soil culture experiments fail to 

 agree. Consecpu'utly, the utmost caution should be 

 observed in drawing conclusions in regard to the 

 causes of soil fertility or infertility, from solution cul- 

 ture studies alone. 



General Summary of All Experi:sients. 



1. The first year's work is substantiated by the data 

 obtained the second year. 



2. Soils verv deficient in nitrogen cannot be much 

 benefited by the addition of lime, phosphorus and 

 polassiuuL Nitrogen alone is much more effective 

 than a combination of these three elements. 



.}. The addition of large amounts of coumarin and 

 vanillin depresses the yield of oats, when the oats are 

 planted immediately after the application of these 

 compounds. 



4. Fertilizers do not prevent the bad effects of cou- 

 marin and vanillin. 



."). Lime, carbon black, and pyrogallol are of little 

 or no l)enefit to plants, when they are added to soils 

 used in these experiments. 



6. Neither pyrogallol nor carbon black increased 

 the effectiveness of fertilizers. 



7: Pyridine and (juinoline, both of which are nitro- 

 genone compounds, have been found to be highly bene- 

 ficial to both oats and corn in all soils tested. 



