28 



The results, therefore, show that the residue left after the 

 removal of the first crop is capable of producing an in- 

 crease of about 8f bushels per acre when ammonia salts 

 are used, and of 2 bushels with nitrate of soda. 



The difference between the Rothamsted and the Woburn 

 results is difficult to account for; but it may perhaps be 

 explained by the difference of soil, and by the fact that at 

 Woburn minerals were applied every year, but at Rotham- 

 sted only in alternate years. The shorter duration of the 

 Woburn experiments may also influence the result ; for at 

 Rothamsted, as we have noticed, at first the residues of 

 ammonia salts had some effect. 



It may be safely concluded, however, that a considerable 

 residue from ammonia salts may remain in the soil after 

 the removal of the first crop. It is probable that much of 

 this is in the form of nitrogenous organic matter, such 

 as roots and stubble left by the crop, which gradually 

 decays and undergoes nitrification. A part, however, 

 seems in the Woburn experiments to have been left 

 untouched by the first crop, and to have remained 

 in its original state in the soil ; for the residue left 

 by the ammonia salts gives its greatest effect in the 

 year following an under-average crop obtained with a 

 direct dressing of nitrogenous manure. For instance, the 

 following results were obtained in the wheat experiments 

 at Woburn in 1895 an< ^ 1896, the former of which years was 

 exceptionally dry, and therefore prevented the full action 

 of the ammonia salts. It will be remembered that mineral 

 manures were applied every year to each plot. 



TABLE XIII. 



In 1895, ^ will b e seen that the application of ammonia 

 salts on plot SA gave practically no larger return than the 

 mere residue from the previous year's manuring on plot SB. 

 But the ammonia salts which failed to act in that year 

 gave an extra return in 1896 ; the yield on plot SA in the 

 second year being within about 3^ bushels of plot SB, to 

 which in 1896 ammonia salts were directly applied. We 

 see from this that, if ammonia salts are not used by the 

 crop to which they are applied, a considerable proportion 

 may remain in the soil for the use of succeeding crops. 

 Moreover, this residue must be to a great extent in its 

 original state ; for if it consisted only of roots and other 

 plant refuse, we should find that when the nitrogenous 



