WATER — SOILS. 



231 



metliod, using standard .lolutionsof potassium nitrate. Total solu])lo salts worodoter- 

 mined by electrical resistance, accordiujz; to Whitney's method. A ta])le is given 

 showing the amount of solul)le salts for different readings of the instrument calcu- 

 lated from a solution curve determined for these investigations. Duplicate determi- 

 nations of nitric nitrogen by the method described showed quite satisfa(!tory 

 agreement, but it was found necessary to protect the solutions from denitrification 

 by the use of formalin and to make the determinations as promptly as possible. It 

 was found "that solutions to which formalin had been added lost none of their 

 nitrates in 4 days, while duplicate solutions to which no formalin had been added 

 lost from 12.2 to 71.6 per cent of their nitrates in that time. . . . This protection of 

 the formalin against denitrification is not permanent and in some cases samples rich 

 in nitrates have lost the whole on standing 5 weeks. In other cases we have observed 

 notable losses on standing 2 to 3 days, this being greater than 50 per cent as a mean 

 of 5 cases." 



As stated above, the electrical method for total soluble salts seemed to be subject 

 to an error due to the absorptive power of soils for soluble salts. In 2 series of 

 experiments in which soil was made into a paste (1) with distilled water containing 

 100 parts per million of potassium nitrate, and (2) with drain water containing 195 

 parts of soluble salts per million, the results were as follows: 



Amounts of soluble salts uctnally present and found hij electrical metliod. 



"These results indicate that the making of the soil paste with dilute soil solutions 

 resulted in fixing some of those salts in an insoluble form. Another set of observa- 

 tions in clearing turbid solutions gave similar results both with nitrates and with 

 other salts, wdiich makes it appear that breaking down the soil texture will throw 

 out of solution solul)le salts already contained in them." 



Nitrification in northwest soils, F. T. Srutt ( Canada Expt. Farms Rpts. 1900, 

 pp. 159-161). — The progress of nitrification was studied in the soils used in the observa- 

 tions on soil moisture referred to on page 233. ' * The method adopted was to weigh out 

 100 gm. of the fresh soil and add thereto 1,000 cc. of ammonia-free distilled water 

 and shake the mixture well for 1 hour. It was then allowed to settle for 1 hour and 

 the free ammonia in an aliquot part at once determined. A further quantitj' was at 

 the same time set aside in contact with a zinc-copper couple (by means of which 

 nitrates are reduced to ammonia) and at the expiration of 24 hours distilled. From 

 the free ammonia in the distillate the amount previously found deducte<l and the 

 remainder calculated to nitrogen, and recorded as nitrogen in nitrates in one million 

 parts of the water-free soil." The results are set forth in the subjoined table: 



Nitrates and nitrites in soils <d different dates. 

 [Parts per million of water-free soil.] 



Brandon soil. 



Date. 



In fallow, 

 1900. In 

 crop, 1899. 



1900. 



May 11 



June 11 



July 11 



Aug. 11 



Sept. 11 



Oct. 11 



Nov. 11 



10. 62 

 15. 21 

 10.99 

 17.94 

 10.67 

 4.55 

 2.53 



In crop, 

 1900. In fal- 

 low, 1899. 



11.45 

 28.20 

 7.65 

 8.42 

 5.51 

 7.91 

 6.40 



Date. 



1900. 



Mays 



Junes 



Julys 



Aug.S 



Sept. 8 



Oct. 8 



Nov.8 



Indian Head .soil. 



In fallow, 

 1900. In 

 crop, 1899. 



3.37 

 6.93 

 22. ;i0 

 22. 70 

 16.71 

 12.20 

 3.99 



In crop, 

 1900. In fal- 

 low, 1899. 



16.22 

 25.70 

 20.00 

 17.20 

 7.20 

 7.32 

 3.97 



