30 EXPERIMENT STATION RECORD. 



cultivated once each week. Apparently the methods emplo3^ed were 

 not sufficiently delicate to show with certainty whether either the 

 frequency or depth of cultivation made any marked difference in the 

 amoxmt of available plant food. 



"The mean change in the sohible salt content in the cuhivated fallow plats, as 

 indicated by the Whitney method, between May 24 and September 15, was 361.1 

 lbs. gain, while the increase on the fallow plat not cultivated was 692.2 lbs. per acre. 

 The mean gain in nitric acid (HNO3) was 830.56 lbs. per acre on the cultivated plats 

 and 371.39 on the ground not cultivated up to August 22." 



Similar observations made May 12 and 13 and August 7 oit plats on 

 which crops were growing showed that the soluble salts remained 

 nearly constant in the upper -1 ft. of soil, there being a tendency to 

 decrease on the whole rather than to increase. The average decrease 

 observed was 174.2 lbs. per acre, while on the fallow plats the mean 

 gain during the period of observations was, as stated above, 397.1 lbs. 



The value of determinations of soluble salts as an index of evapo- 

 ration from the soil is discussed at some length. 



The electrical resistance by Whitney's method, computed as sodium 

 chlorid, the amount of nominal alkalis, partly and p()ssi>)ly largely 

 sodium carbonate, as indicated by Ililgard's method for detecting black 

 alkali, and the amount of nitric acid were determined August U in 

 plats of productive and improductive humus soils planted to potatoes 

 and treated with sodium and potassium car})onates and Avood ashes. 



"The method of detecting the nominal alkalis in the soil whicli lias been used 

 consisted in weighing into a small muslin sack 50 gm. of the fre.sh soil and wash- 

 ing this during 2 minutes with 255 cc. of distilled water poured into the sack in a 

 mortar. By holding tlie sack closed and on its side and working it with the pestle, 

 turning it from time to time, the soluble salts are quickly taken up by the water. 

 The water is then wrung from the sack and the solution poured into a mug to settle. 

 When clear 25 cc. of the filtered solution is evaporated to dryness and then redis- 

 solved and titrated against deci-normal hydrochloric acid. The balance of the soil 

 sample is dried to determine the water content and the percentage obtained used in 

 calculating the alkalinity of the soil. . . 



"In [determining nitric acid] a weighed fresh sample [of soil], usually 50 gm., 

 was ])laced in a small nmslin sack in a mortar. Into the sack was poured 250 cc. of 

 distilled water. Holding the sack closed in one hand and the pestle in the other the 

 soil was worked by pestling and turning during 2 minutes, when the sack was removed 

 an<l drained by wringing and squeezing. 



"The turl)id solution was transferred to mugs and allowe<l to stand for from (i to 

 12 hours to settle. It was found, however, that long standing was not i>eriiiissil)le, 

 especially with humus soils, on account of a tendency to denitrification. 



"Corrections were made, in the calculations, for the moisture in the fresh sample 

 used by determining the water content in the balance of the sample. This plan was 

 followed to avoid the danger of increasing the nitrogen content by drying and to 

 avoid changing the solubility of the soil by heating, our object being to get the total 

 soluble salts l)y the Whitney method and the amount of nitric; acid by the phenyl 

 sulphate m('tho<l of Ticffmann and Beam." 



The largest amount of nominal alkalis was found in the surface 6 in. 

 of the soil treated with sodium carlxniate. In general the amount of 



