22 



THE RELATION BETWEEN QUANTITY OF FERTILISER 

 AND CROP YIELD. 



These investigations started from the Broadbalk result that the 

 second increment of nitrogenous fertiliser produced a larger incre- 

 ment of yield than the first. If this proved generally true in farm 

 practice it would mean that under normal conditions of price a 

 farmer would be well-advised to manure pretty liberally. The 

 Broadbalk experiment has, however, certain unpractical features, 

 and a series of field trials under ordinary farm conditions has been 

 carried out. 



The results with wheat in 1920 favoured this view (Report 

 1918-20, p. 79), the yields without nitrogen being 28.9 bushels 

 and with the higher dressing 35.9 bushels per acre. Unfortunately 

 both in 1921 and 1922 the wheat crops were very poor, the yields 

 without nitrogen averaging 17.5 and 13.4 bushels per acre respec- 

 tively, which values were hardly raised in 1921, and only to 17.1 

 and 19.7 bushels by the single and double dressing respectively in 

 1922 (p. 93) . No definite conclusion can be drawn from these 

 figures. 



Potatoes made much better growth. The tops were not 

 weighed, but the tubers increased in yield with successive incre- 

 ments of sulphate of ammonia, and gave a record crop for this 

 land. The increases for the second increment, however, were not 

 greater than for the first, but probably slightly less ; nevertheless 

 under ordinary conditions of price the results would have been 

 very profitable. The figures were : — 



GREAT HARPENDEN FIELD: POTATOES, 

 (Mean of duplicate set.) 



1922. 



Uasal manure (with duni<) equals 4 cwt. super, 1'/^ cwt. sulphate/potash. 



Masai manure (no dung) equals 6 cwt. super, 2 cwt. sulphatc/potasli. 



(1) Of this A'/a cwt., 3 were applied with the seed, and I'/g given later as 



top dressing. 



These apparent discrepancies are being fully gone into during 

 ihe coming season. 



THE SOIL POPULATION AND THE PRODUCTION OF 

 PLANT FOOD IN THE SOIL. 

 The important investigations by Mr. Cutler and the staff of the 

 I'rotozoological Department have necessitated considerable revi- 

 sion of our ideas of the soil population. It had always been 

 supposed that the numbers of organisms present in natural soil 



