274 University of California Publications in Agricultural Sciences [Vol. 3 



barnyard manure with the Berkeley clay adobe soil. The applications 

 of CaC0 3 and of CaS0 4 were made on March 9, 1917, and in the 

 manner and quantities indicated in the tables. Control soils, untreated 

 with either lime or gypsum, were, of course, included in the experi- 

 ment, but were otherwise treated like the other soils. The soils in 

 all pots were sampled three times, at considerable intervals, as shown 

 in the tables. The samples were taken so as to represent the whole 

 depth of the soil layer in the pot and of different parts thereof. Eight 

 hundred gram portions of these samples, in air-dry condition, were 

 mixed with 1600 cc. of distilled water in large bottles and allowed to 

 digest for six days with occasional shaking during every day. After 

 six days, the solutions were filtered through Pasteur-Chamberland 

 pressure filters and analyzed by gravimetric or volumetric methods 

 for the constituents named in the tables. Large enough aliquots could 

 be employed, owing to the method which we have devised and described 

 above for mixing the soil and water, to insure accurate results by the 

 standard methods of analysis intended for larger quantities of the 

 same substances. We employed no checking system by means of 

 germinating plants, such as that used by Briggs, because (1) we do 

 not believe that a few days' growth of plants constitutes any reliable 

 criterion regarding any factor in plant growth, and (2) the evidence 

 obtained by Burd, Hoagland, and Stewart has demonstrated that for 

 plants grown to maturity, an intimate relation holds between the 

 nature of the soil solution and absorption of nutrients by plants grow- 

 ing in such soil solutions in every stage of growth. In other words. 

 we contented ourselves with trying to determine whether or not the 

 soil solution is enriched with respect to potassium and other elements 

 by the treatment of soil with CaCO, or with CaS0 4 as indicated by 

 tin com position of the soil extracts obtained by us. The results of the 

 analyses of the soil extracts are shown in the subjoined tables, in which 

 the amount of every ion sought and found in the solution is expressed 

 in parts per million of the soil. 



For the sake of greater simplicity and brevity, we shall at first 

 discuss the tables separately. 



In table 1, we see the results obtained with the Oakley soil, from 

 which it is clear that both lime and gypsum are without effect on the 

 amount of water soluble potassium in that soil. The latter behaves in 

 this respect like the Oatman soil from Riverside County in this state, 

 which Hii-_!<4s and Breazeale have studied. There is little reason to 

 believe, likewise, from the data under consideration that the water 



