EXPERIMENTS AT CIRENCESTER 67 



up by the plant. If therefore no nitrate is assimilated after a 

 certain stage of growth is reached by the crop, it is the nitrate 

 of soda of which the smallest proportion will remain unused, 

 while the nitrate from the ammonia salts may be to a more 

 considerable extent left unconsumed. 



The larger residue from the ammonia salts left by barley 

 than by wheat in the Woburn experiments is doubtless partly 

 due to the somewhat later date at which the top-dressing of 

 ammonia is applied in the case of barley, but it is also probably 

 in part due to the much greater root development of the autumn 

 sown wheat plant at the time when the top-dressing of ammonia 

 salt is applied, and to the consequently more rapid powers of 

 assimilation of the wheat crop. 



The valuers of the unexhausted manures left in the land by 

 an out-going tenant are not in the habit of attaching any money 

 value to a supposed residue of ammonia salts, or nitrate of soda, 

 when these have been applied to a preceding cereal crop. The 

 facts before us show, however, in an unmistakable manner, that 

 when a liberal application of ammonia salt is applied late to the 

 land, and especially to a soil poor in lime, a considerable residue 

 may sometimes, and notably in a dry season, remain in the soil 

 after harvest. 



4. Experiments at Cirencester At the Royal Agri- 

 cultural College, Cirencester, experiments on the continuous 

 growth of barley with various manures were conducted for 

 seven years, 1885-91. The soil was a rather calcareous loam, 

 irregularly mixed with clay, and belonged to the oolite for- 

 mation. Analysis showed that both soil and subsoil were well 

 provided with nitrogen, potash, and phosphoric acid. In these 

 experiments the farmyard manure was ploughed in in the spring. 

 The superphosphate and kainite were harrowed in before 

 sowing the seed. The sulphate of ammonia was applied as a 

 top-dressing to the growing crop about the middle of May. 



In the following table the average produce during six years, 

 1885-90, is given ; the crop of 1891 is omitted, having 

 suffered much from weeds and birds. Each produce mentioned 

 in the table is the mean of two plots. 



