298 SCIENCE PROGRESS 



a controlled supply of water; it implies not only irrigation, but 

 drainage. 



The remainder of this article will therefore be still further 

 restricted in scope, not merely to the best cottons, but to cotton 

 grown under irrigation. 



The primary need which confronts any person concerned 

 with the development of cotton-growing in a new district, is 

 the need for knowing how the plants behave under optimum 

 conditions. This statement is a very commonplace one, but 

 the knowledge is practically unobtainable at present. We 

 have first to ascertain the optimum conditions in general 

 terms, and then to ascertain how the plants react to them. 



A part of this question also concerns the grower of cotton in 

 an established area. The optimum conditions have presumably 

 been found there by trial and error, but economic botany has 

 rarely effected such an analysis as shows exactly how the 

 plants behave. 



Lastly, there is the purely scientific inquiry into the causes 

 of such behaviour. The economic investigator should carry his 

 analysis of existing conditions down to the point at which they 

 resolve into general problems. To study the physiology of 

 growth on the lint-hair cells of cotton is not likely often to be 

 practicable ; it is quite practicable to study the growth of lint- 

 hair cells under existing field conditions sufficiently minutely so 

 that results obtained by the study of a fungus in another con- 

 tinent can be directly utilised to interpret the phenomena 

 observed in cotton. 



Taking the two economic issues only, viz. determination of 

 the optimum conditions, and recognition of the reactions of the 

 plant to those conditions, there is every prospect of a great 

 and rapid advance in the application of science to cotton- 

 growing. 



As regards the determination of optimum conditions. The 

 development of statistical methods, with the concurrent 

 analysis of the errors to which experimental plots are subject, 

 the devise of small plots and systematic scatter, and the use 

 of graphic methods, have combined to make it practicable to 

 determine the optimum conditions of cultivation for a given 

 year on a small area, and at small cost. Whether repetition 

 on the same scale is necessary for several years will depend 



