SOILS — FERTILIZERS. 619 



and rubber tnblng to the manometer and the soil container. The flow of water 

 from the reservoir to the soil container is regulated by stopcocks and the 

 capillary lift of the soil is indicated by the rise of the mercury in the 

 manometer. 



Losses of moisture and plant food by percolation, G. S. Fbaps (Texas Sta. 

 Bid. 171 {lOl^t), pp. 5-51). — The results of three years' experiments made to 

 ascertain the amount of percolation and evaporation from eight different soil 

 types of Texas and the effect of cultivation, manure, and fertilizers upon the 

 amount of water percolating and on the losses of plant food therein are re- 

 ported. The soils used were Norfolk sand, Orangeburg fine sandy loam, Hous- 

 ton loam, Houston black clay, Yazoo clay, Miller fine sandy loam, Crawford 

 clay, and Lufkin fine sandy loam. 



Percolation was found to be greater through uncultivated clays and loams 

 than through uncultivated sands and sandy loams. Cultivation increased per- 

 colation through the latter, but had little effect upon the former. Potassium 

 sulphate and manure increased percolation through sandy soils but the former 

 decreased percolation through clay soils. The fall application of manure was 

 more effective in increasing percolation than the spring application. 



The amount of nitrates in water percolating from uncultivated soils was 

 related to a certain extent to the total nitrogen of soil and subsoil. More 

 nitrate appeared in the percolates from manure applied in the fall than in the 

 spring. An application of sodium nitrate increased the nitrates in the succeed- 

 ing percolates of two of the soils, but with the other six no effect was observed 

 until three or four weeks later. Only small quantities of potash appeared in 

 the percolates from most of the soils, even after heavy applications of potash 

 were made, the maximum loss being 12 per cent in three years from the Norfolk 

 sand and the losses from the other soils varying from to 4.5 per cent. From 

 9.7 to 66.6 lbs. of potash per acre per year were lost by percolation from un- 

 cropped, uncultivated, and unfertilized soils, and the losses were to a certain 

 extent related to the active potash of the soil. The losses of phosphoric acid in 

 the percolates were very small, but the losses of lime were large, varying from 

 70 to 5S2 lbs. per acre from the uncropped, uncultivated soils. The lime losses 

 were in a general way related to the quantity of lime soluble in strong hydro- 

 chloric acid. 



Effect of moisture content of a sandy soil on its nitrog-en fixing power, 

 C. B. LiPMAN and L. T. Shaep (Bot. Gaz., 59 (1915), No. 5, pp. 402-406).— 

 Studies of the natural nitrogen fixing flora of a light sandy soil from a walnut 

 grove with reference to moisture content are reported. The soil-culture method 

 was used. 



It was found that nitrogen fixation in this soil by means of its natural flora 

 and under optimum temperature conditions takes place most actively with a 

 water content varying from 20 to 24 per cent, based on the air-dry weight of 

 soil. "With 28 per cent of moisture there was a slight decline in nitrogen fixa- 

 tion. With 32 per cent a marked decrease in nitrogen fixing power of the soil 

 was evident, and a still gi-eater decrease was noted with 36 per cent. Almost 

 no nitrogen fixation, or very little, took place with a moisture content of 4 per 

 cent (air-dry basis), but a very marked increase occurred when 8 per cent was 

 present, and amounts of moisture equivalent to 12 per cent (air-dry basis) gave 

 about the same nitrogen fixation as 32 per cent moisture. 



It is concluded from these results that for the soil used the aerobic forms of 

 nitrogen-fixing bacteria do best with a 20 per cent moisture content. At higher 

 percentages of moisture up to 24 per cent the anaerobic forms become much 

 more active, while the aerobic forms are depressed in their nitrogen-fixing 

 powers. 



