378 INVESTIGATION BY CULTURE EXPERIMENTS 



annual rainfall at Rothamsted, and of this amount 50 per cent passes 

 off in drains, at a depth of 40 inches, and 50 per cent is evaporated, 

 from a soil kept free of vegetation. Roughly, the evaporation from 

 a bare soil may be regarded as a constant, to be subtracted from 

 the rainfall to find the drainage (and run-off, if any). Thus, if we 

 regard 14.25 inches as the constant for evaporation at Rothamsted, 

 the drainage should be 6.24 inches for 1898 and 24.44 inches for 

 1903, while the actual records show 7.90 and 23.59 inches, respec- 

 tively. 



Of course the evaporation can be markedly reduced by culti- 

 vating the surface as soon as practicable after each rain, in order 

 to destroy the capillary connection and to maintain a dust mulch, 

 and thus largely preventing the rise of moisture to the surface. 

 On the other hand, evaporation is greatly increased by growing 

 crops, so that during the growing season the drainage would be 

 less on the ordinary field than from the bare soil. 1 



BARLEY EVERY YEAR ON Hoos FIELD, ROTHAMSTED 



Table 66 presents in summarized form the data secured from 

 Hoos field, where barley has been grown every year since 1852. 

 These experiments help to answer some important questions con- 

 cerning which neither Agdell nor Broadbalk give any information. 

 The yields, as an average of 55 years, vary from 14.8 bushels on 

 the unfertilized land, and 15.7 bushels where only the sulfates 

 of potassium, magnesium, and sodium were used, to 43.9 bushels 

 with sodium nitrate and acid phosphate, and 47.7 bushels with 

 farm manure (15.7 tons a year). 



As an average of the 3<D-year and 25-year periods, the yields 

 have decreased nearly 10 bushels per acre on all plots receiving 

 nitrogen, undoubtedly because the 43 pounds of nitrogen was not 

 sufficient for larger crops, after deducting losses by leaching. It 

 will be remembered that the second addition of 43 pounds of nitro- 



1 Ingle reports some computations in his "Manual of Agricultural Chemistry," 

 page 76, in which the drainage is reckoned at about 86 inches; but probably the 

 intention was to use 8.6 inches, which would reduce his estimated "enormous loss 

 of phosphoric acid" to a very insignificant amount quite in harmony with other 

 data, such as he gives on page 77. 



