858 Journal of Agricultural Research vol. vr, no. 22 



to differ from one given in terms of the salt recoverable from the same 

 soil. In actual field practice salt lands must be classified in terms of 

 the amount of salt recoverable from them and not in terms of the amount 

 which has been added to them, which is not ascertainable. 



PLAN OF THE EXPERIMENTS 



The general plan of all of the experiments was as follows: Ordinary 

 drinking glasses were filled with 300 gm. of air-dry soil. The salts were 

 added from a stock solution of known strength and ranged in amount 

 from nothing up to concentrations sufficiently strong to prevent plant 

 growth entirely. Distilled water was added to each glass to moisten the 

 soil thoroughly. Seven seeds of wheat (Triticum spp.) were planted in 

 each glass, and after germination the number of plants was reduced to 

 five if more than that number came up. Bluestem wheat was used in 

 1913 and 1914, and Marquis wheat in 1915. To prevent loss of moisture, 

 the glasses were covered with glass plates until the plants emerged. 

 After the emergence of the plants, the pots were weighed daily, and by 

 the addition of distilled water the moisture content was brought back to 

 the original condition. In 191 3 and 1914 the experiments were con- 

 ducted in triplicate, while in 191 5 they were in duplicate. 



The wheat was allowed to grow from 15 to 18 days, when the plants 

 were cut at the surface of the ground and weighed immediately in a 

 closed tube. After cutting the plants the soils from each series of glasses 

 were mixed, dried, and analyzed for water-soluble salts. 



CONVERSION OF CARBONATES 



In the experiments where sodium carbonate was added to the soil, the 

 analysis of the soil after the wheat had been cut brought out the fact 

 that a portion of the sodium carbonate that had been added was not 

 recoverable. 



When only a small quantity of sodium carbonate was added, none 

 could be recovered at the end of the experiment, but the quantity of 

 sodium bicarbonate was greater than in the untreated soil. With the 

 addition of larger quantities of sodium carbonate both salts were recov- 

 ered at the end of the experiment, but their sum was always less than the 

 quantity added at the beginning. 



It is apparent that a portion of the sodium carbonate added to the soil 

 was converted into sodium bicarbonate. In order to determine what 

 proportion of the original quantity of sodium carbonate could be ac- 

 counted for at the end of the experiment, it was necessary to add together 

 the quantity of sodium carbonate recovered as such and the quantity 

 represented in the form of sodium bicarbonate. 



The conversion of sodium carbonate to sodium bicarbonate results in 

 an increase in weight of the salt at the ratio of 44 to 70 — that is, the 



