60 



EXPERIMENTS UPON WHEAT 



marked in a wet than in a dry season, is confirmed by a 

 further examination of the records over a series of years. 

 Taking the last thirty years and dividing them into two groups 

 according as the rainfall is above or below the average, and 

 then comparing the yields of the two plots, which receive equal 

 amounts of nitrogen, but one as nitrate of soda and the other 

 as ammonium-salts, we find that in the dry seasons the yield 

 from ammonium-salts is 86 '6 per cent, of the yield from 

 nitrate of soda. In the group of wet seasons, however, the 

 yield from ammonium-salts is only 78 "8 per cent, of that given 

 by nitrate of soda, as shown in Table XXII. Thus the wet 

 seasons are on the whole more favourable to nitrate of soda 

 than to ammonium-salts. Presumably in the very wet and 

 cold seasons the conditions are unfavourable to the nitrifica- 

 tion of the ammonium-salts, and the immediately available 

 nitrate of soda is more effective. 



Table XXI 1. — Broadhalk Wheat. Comparison of the yield of Dressed 

 Grain ivith Nitrogen as Ammonium-salts or Nitrate of Soda, in seasons 

 when the Rainfall was heloio or above the average, 30 yeai^s (1873-1902.) 



One of the most critical periods in determining the yield of 

 wheat appears to be the winter months ; if the wheat be sown 

 in October or early November it spends the next three or four 

 months almost wholly in developing its system of roots. 

 Should the weather be wet and the soil in a satui'ated condition 

 the root- system will be restricted, both because of the deficient 

 aeration and because the roots need not extend far in order to 

 obtain the water necessary for its growth. From the indifferent 



