20' 



alkali by interaction with carbonate of lime in the soil. The re- 

 medy is to apply such manures in the winter or in conjunction 

 with superphosphate. 



E. J. RUSSELL. The nitrogen problem in crop production. 



(Journ. Roy. Agr. Soc. England, 69, 1908, Pp. 104-114) Exper. 

 Stat. Rec. 9 Vol. XXII, No. 3, Febr. 1910, Washington. 



This article is based mainly upon investigations made at Ro- 

 thamsted and discusses the sources of nitrogen available to the 

 farmer, including nitrogen compounds in the soil, purchased ma- 

 nures, purchased feeding stuffs, nitrogen fixed by bacteria alone 

 and in symbiosis with leguminous plants, and losses of nitrogen on 

 the farm due to drainage and to bacterial action. The value of a 

 rotation adapted to the maintenance of the nitrogen supply is 

 pointed out. 



Some recent Agricultural Field Trials in Great Britain. - 



Nature, Vol, 83, p, 313. London, May 12, 1910. 



A Review of the following articles: 



" West of Scotland Agricultural College, Reports on experiments; 

 Durham County Council Education Committee, Report of Dairy 

 Investigations ; Northumberland Education Committee, Guide to Ex- 

 periments for 1909; East Sussex Education Committee, Experiments 

 on Meadow Hay; Agricultural Students' Gazette, Experiments on 

 Pasture; Journal of the Department of Agriculture and Technical 

 Instruction for Ireland. 



" Field experiments are perhaps the most difficult of all expe- 

 riments to interpret. The growth of plants is influenced by so 

 many factors temperature, water supply, food supply, etc. that 

 small variations in the condition may lead to marked differences 

 is result. The soil is a very complex mixture, and not uncommonly^ 

 shows some variation even within the limits of a single field. Even 

 if the soil itself is tolerably uniform a slope may introduce irre- 

 gularities. The influence of a tree or hedge makes itself felt for 

 some distance, while the effect of the previous treatment of the 

 ground if often very great. 



" In the experiments studied in the above articles in no case 

 is the experimental error indicated, nor can we find it has been 



