1918J SOILS FERTILIZERS. 211 



" A soil containing the amount of moisture corresponding to its hygroscopic 

 coefficient loses water very rapidly when exposed to an ordinarily dry atmos- 

 phere, but in determining the hygroscopic coefficient the time necessary to 

 transfer the soils from the absorption boxes to weighing bottles is so brief that 

 the loss during the transfer is too small to affect appreciably the accuracy of 

 the results. 



" Hllgard's method for the determination of the hygroscopic coefficient, car- 

 ried out exactly as he described it, gives reliable results. However, the loose 

 sheets of glazed paper thus involved are very inconvenient when many deter- 

 minations are to be made and may advantageously be replaced by shallow 

 trays, either of aluminum or of copper. . . . Any considerable increase in the 

 size of the absorption boxes over that recommended by Hllgard or the use of a 

 larger number of exposed samples within the boxes of the same size cause too 

 low results, unless the time of exposure be greatly increased." 



A bibliography of 27 titles is appended. 



Some factors affecting nitrate-nitrogen accumulation in soil, P. L. Gainey 

 and L. F. Metzlek ([/. S. Dept. Agr., Jour. Agr. Research, 11 (1911), No. 2, pp. 

 .'i3-64). — The authors report the results of extensive investigations at the 

 Kansas Experiment Station to ascertain the influence upon nitrate-nitrogen 

 accumulation in soils of variations in some of the more important factors con- 

 trolling aeration. 



In preliminary experiments, variations in the quantity oi- soil amounting 

 to from 50 to 1,000 gm. had little if any effect upon nitrification. Variations 

 in the depth of columns of loose soil of from 0.25 to 20 in. did not produce 

 appreciable differences in nitrification provided the soil Avas left loose, and 

 nitrification was apparently no less vigorous 20 in. below the surface than at 

 the surface. Packing the soil in a thin layer was without effect, but upon in- 

 creasing the depth of column packing (reducing the volume from 14 to 9) 

 resulted in a marked decrease in nitrate accumulation, the latter becoming 

 negative only a few inches below the surface. A decrease in the ratio of 

 surface exposed per 100 gm. of soil of from 314 sq. cm. to 2 sq. cm. had no 

 effect upon nitrification. The shape and size of the container and methods 

 of preventing evaporation and contamination were without effect except when 

 the container was tightly stoppered and when the volume of inclosed air was 

 relatively small in proportion to the soil volume. 



These observations led to more detailed experiments of the effect upon 

 nitrification of variations in depth of column and compactness of soil ; of depth 

 of column, moisture content, and compactness ; of soil in sealed containers as 

 compared with a soil surface exposed to the atmosphere; and of unbroken soil 

 columns as compared with broken columns. The data are tabulated, discussed 

 in some detail, and available experimental data reported by other investigators 

 relative to the influence of different degrees of aeration upon nitrate formation 

 briefly reviewed. 



The authors conclude that " as the moisture content of a soil decreases, in- 

 creasing the compactness from a very loose condition will increase the accumula- 

 tion of nitrate nitrogen. With any degree of compactness tested the optimum 

 moisture content will be reached when the soil contains approximately two- 

 thirds the total amount of moisture it will retain. Aeration will be sufficient 

 to the depth of 1 ft. with any degree of compactness, provided the moisture con- 

 tent does not exceed the above relation. 



" Increasing the depth of column up to 2 ft. does not, as far as tested, alter 

 the above relations. In fact, the accumulation of nitrate nitrogen increases 

 with increasing depth down to 2 ft., so long as the moisture does not exceed 

 approximately two-thirds saturation. Nitrate nitrogen accumulates more 



