if it be a rule, to plot 5 Drake field may explain the low 

 yield of cotton there, a deficiency both of potash and of 

 nitrogen being manifest. Likewise in plot 7, Station gar- 

 den, Ave find a large decrease in the percentages of all three 

 constituents, although two of them have oeen added to the 

 soil, and here, also, we find a relatively low yield. 



In connection with this work, it has been thought well to 

 make a complete analysis of the two soils. In view of the 

 fact of their similarity geologically, both being classed as 

 light sandy soils, and the additional fact that one is very 

 poor and the other rich, a comparison of their chemical com- 

 position will be interesting. 



CHEMICAL ANALYSIS OF SOILS. 



Moisture 



Insoluble Silica. . 

 Soluble Silica. . . . 



Alumina 



Oxide Iron 



Lime 



Magnesia 



Soda 



Potash 



Phosphoric Acid. 



Nitrogen 



Organic Matter . . 



Humus . . . 



Available Inorg. Matter. . 



Humus Silica '. . 



Humus Phosphoric Auid. 



As will be observed, both soils have a high percentage of 

 insoluble silica, that of the field exceeding that of the gar- 

 den nearly two per cent. Oxide of iron in the hydrated 

 condition is believed by some to increase in soils the absorp- 

 tive power of gases, and particularly, of moisture. Both 

 of our soils are low in this constituent, with the advantage 

 in favor of the garden. Estimated in terms of the poorer 

 soil, the garden soil is 28.6 % higher in oxide of iron than 

 the other. If the minimum limit assigned to lime in light 

 sandy soils by writers on this subject be correct, both 

 of these have a sufiiciency of this valuable constituent, 

 the garden having 40.5 % more than the field. In both 

 potash and phosphoric acid, on the other hand, the gar- 

 den soil is poorer, about 1 "u in the former and 26.4 "o in 

 latter. What has just been said applies to total phos- 



