19 



which a little calcium carbonate was added j)roduced 

 nearly as good peas and sorghum as did the extract 

 which had been reduced and made to volume with 

 tap water, or ashed, and made to volume in a similar 

 manner. The distillate of the soil extract w^as slightly 

 better as a medinum for growth than distilled water 

 made in the same glass still used in distilling the soil 

 extract. 



Evidence that the amount of manganese found in the 

 several soil extracts used was sufficient to cause great 

 injury to plants is here given. To the distillate of a 

 toxic soil extract wdth a high manganese content, man- 

 ganese sulphate w^as added to equal the manganese 

 content of the original extract. Both roots and 

 tops of plants died in this culture. In two 

 other instances, soil extracts containing but traces 

 of manganese, and capable of supporting a very 

 satisfactory growth of sorghum, have been made 

 extremely toxic through the addition of manga- 

 nece sulphate at the rate of 200 p. p. m. These 

 results are not surprising in view of the fact tliat Mc- 

 Cool (12) has shown that a solution of N/4000 MnCL 

 (18 p. p. m.) is injurious, and N/2000 MnCL (36 p. p. 

 m.) prevents root growth of pea seedlings, using dis- 

 tilled water as the solvent. 



One of the most interesting experiments tending to 

 show the inorganic nature of the toxic body in our soil 

 extracts is given in Table VI. The soil extract used 

 contained 15.8 p. p. m. of aluminum, and 80 p. p. m. of 

 manganese. The cultures grew from Dec. 8th to Dec. 

 29th, 1917. 



The culture solution to which 3 cc. of N/1 NaOH 

 were added was slightly alkaline to litmus and to phe- 

 nolphthalein; but those to which less than 3 cc. w^ere 

 added, gave no reaction with these indicators. The 

 cultures which received lime were treated with an ex- 

 cess of CaO, and vigorously shaken for about 20 min- 

 utes, after which the solution was left undisturbed for 

 several hours. The excess calcium hydroxide was 

 then carbonated. In all of the previous experiments, 

 the compounds precipitated by alkaline hydroxides 

 were filtered out; but in this case, the precipitate was 

 left in the culture medium. Special attention is call- 

 ed to the fact that partial precipitation proved very 

 much more effective in reducing the toxicitv of this ex- 

 tract than various degrees of dilution. In fact, the 



