Ixxvi Department of Soil Investigation 



duplicate or triplicate plats arranged in series. The agitation of 

 this subject among station workers promises to bring about im- 

 provement in the conduct of field experiments, which experiments 

 have been for years among the least accurate of the forms of ex- 

 periment station activity. 



The concrete tanks for soil experiments have been completed and 

 are now in use. They are intended to furnish receptacles for bodies 

 of soil of sufficient size to produce plants in a normal manner under 

 approximately field conditions, and yet afiford opportunity for meas- 

 uring a large number of the factors afit'ecting plant growth. The 

 construction is of concrete, but the tanks will be lined. 



Each tank is four feet two inches square with a maximum ver- 

 tical depth of four feet six inches and a minimum depth of four 

 feet. There are twenty-four tanks placed in two rows of twelve 

 tanks each. Between the rows of tanks is a tunnel, the bottom of 

 which is ten feet below the top of the tanks. The tunnel is six 

 feet wide. From the lowest point in each tank is an outlet tube 

 two inches in diameter and tin lined. It is made large enough to 

 permit of easy cleaning and has no bends in it. A piston runs 

 through the tube to within four inches of the upper end. Between 

 the perforated head of the piston and the soil, glass wool is to be 

 inserted. The piston can be withdrawn if it is desired to clean the 

 tube. 



Drainage water from each tank will be caught in a receptacle 

 in the tunnel. The lining in the tanks will prevent any soluble 

 material in the concrete from appearing in the drainage water. A 

 constant water table at any depth may be maintained by raising the 

 rubber tube leading from the outlet tube to a corresponding point 

 below the surface of the soil in the tank. 



The tanks as described will each contain between three and four 

 tons of soil, and the surface will constitute approximately .0004 of 

 an acre. They are built with special reference to durability, so that 

 it will be possible to plan for experiments to extend over a long 

 period. The quantity of soil contained is not too large to allow of 

 accuracy in sampling and yet is sufficiently large to closely resemble 

 field conditions, which is not true of the quantity contained in pots. 

 No covering is to be placed over the tanks, but in every way nat- 

 ural conditions are to be permitted. The top soil and the subsoil 

 will be placed in their relative positions. It is expected that the 

 rainfall will be sufficient to meet the needs of the crops, but if plants 

 suffer during periods of drought they can be watered artificially. 



