5i6 



Journal of Agricultural Research 



Vol. XV, No. 9 



analyses made of colloid suspensions not here included agree in showing 

 lower iron and higher silica, calcium, and manganese in this colloid 

 suspension than in the other suspensions. The variation in the composi- 

 tion of the other colloid suspensions is probably due to anal3'^tical errors. 

 The low silica obtained on the colloid from the nitrate-treated soil is 

 probably an analytical error, as other determinations on colloid from 

 similarly treated soil gave about the same percentage of silica as obtained 

 on the other colloids. 



Fraps {4) found that the ammonia-soluble inorganic soil colloids varied 

 in composition with the percentage of colloids in the soil. The percentage 

 of silica decreased and the percentage of aluminium increased with in- 

 creasing percentages of total colloid, in soils of different character and 

 from dififerent localities. 



There is a marked difference between the composition of the whole 

 untreated soil and of the inorganic colloid suspensions obtained from it. 

 Aside from the difference in the silica, the differences in the iron and 

 aluminium are evidently significant, especially in the light of determin- 

 ations made on a soil of different physical and mechanical properties. 



A heavy adobe soil from near Whittier, Cal., was treated with different 

 inorganic substances. Determinations of inorganic colloid suspension 

 were made, and the colloid partially analyzed. Table X gives the results 

 obtained. 



Tabids X. — Partial analysis of whole soil, and of inorganic colloid suspensions obtained 

 from the same soil after having received different inorganic treatments. Heavy adobe 

 soil, near Whitney, Cal. 



The analysis of the inorganic colloid suspensions obtained from the 

 adobe soil tmder various treatments does not indicate any fundamental 

 differences in the composition due to these treatments. 



The percentage of iron and aluminium in the colloids is markedly 

 higher than in the original soil, the same relation as was found in the soil 

 from Riverside (Table IX) . The percentage of silica in the original adobe 

 soil is about the same as in the colloid suspensions obtained from it. 



A comparison of Table IX with Table X shows that the percentage of 

 iron and aluminuim in the colloid suspensions obtained from the River- 

 side soil is higher than that in the colloid suspensions obtained from the 



