424 EXPERIMENT STATION EECORD. 



added to the pot. They were added iu solution, . . . but if added after plant- 

 ing, the solution was diluted with about 200 cc. of water. 



" The seed were weighed' out so that each pot received the same amount of 

 seed within 0.1 gm. Water was added to one-half the saturation capacity of 

 the soil. If this quantity was found to be too gi-eat, it was afterwards reduced, 

 but this was the case in only a few instances. The pots were weighed, placed 

 on scales three times a week, and water added to restore the loss in weight. 

 If the plants needed water between these weighings, such quantity was added as 

 appeared necessary. The object of the weighing was to maintain as closely as 

 possible a constant amount of water in the soil. 



"A few of these experiments were conducted in a greenhouse . . . and a 

 number were made on trucks covered with wire mosquito netting. The trucks 

 were pulled into the house when a storm threatened. Later experiments were 

 made in houses covered with canvas. These houses appear to be very well 

 suited to pot experiments under our climatic conditions. They are much better 

 for this purpose than glass houses, for the reason that the circulation of the air 

 is considerably better and the house does not become so heated as a glass house 

 would. Some of these experiments were carried on iu houses with glass roof 

 and canvas sides. This also appears to be a good form of house for our climatic 

 conditions. A house with glass top and wire mosquito-netting sides is also 

 being used. In some respects this is better than the canvas house, but in other 

 respects it is not. The canvas houses are somewhat cooler. The open house is 

 hotter, but the plants are of heavier growth than in the canvas house." 



The author summarizes his I'esults as follows : 



" The plant food withdrawn from the soil by the plant depends upon the 

 form of combination of the plant food, its protection or nonprotection by en- 

 crusting particles, the action of weathering agencies upon it, and the nature of 

 the plant. 



" The composition of the soil extract, by any solvent, depends upon the 

 quantity of the phosphate exposed to the solvent, and its solubility -under the 

 conditions of the extraction, the solubility of the material which protects phos- 

 phates, and the fixing ability of the soil for phosphoric acid from the solvent 

 in question. 



" Fifth-normal nitric acid dissolves phosphates of lime completely, but dis- 

 solves such iron and aluminum phosphates as usually occur in the soil only to 

 a slight extent. It thus distinguishes between these two classes of compounds 

 in the soil. 



" Fifth-normal nitric acid may not distinguish between phosphates which 

 have unequal values to plants. Soils should be compared which probably con- 

 tain the same kinds of phosphates. 



" One per cent citric acid has a lower solvent power for mineral phosphates 

 than fifth-normal nitric acid. The solvent power of other solvents is discussed. 

 Fifth-normal nitric acid is preferred. 



" Soils absorb phosphoric acid in solution in fifth-normal nitric acid and 

 other solvents. 



" The percentage of the added phosphoric acid absoi-bed by the soil increases 

 as rts content of oxids of iron and aluminum increases. 



" Residues from the extraction of the soil with fifth-normal nitric acid and 

 with stronger acids may have nearly as great absorbing power as the original 

 soil. 



"The phosphoric acid absorbed by soils is not extracted by the first extrac- 

 tion with fifth-normal nitric acid, but its effect is evident in the fourth, and 

 sometimes in the sixth, extraction. 



" Natural soils resemble soils which have received potassium phosphate in 

 their behavior to fifth-normal nitric acid in successive extractions. 



