SCIENCE AND COTTON 299 



partly on the climate of the area, and partly on the extent 

 to which the causes of success or failure can be analysed 

 by such methods as those to which we shall next advert. 

 Such a set of small plots would include five of every kind, 

 and would be sown at five different dates, with plants 

 spaced in five different ways, and with water-supply or manures 

 varied similarly according to the practical requirements of the 

 district. Such work is worth doing even in an established area, 

 on account of the insight it gives into local customs. 



As regards the recognition of the reactions of plants to the 

 environment, some simple methods devised by the writer make 

 this much less difficult. A series of plants, varying in elabora- 

 tion of plot-arrangement according to circumstances, are put 

 under daily observation for such features as their rates of 

 growth, flowering, and fruiting. Graphs are constructed from 

 the data thus obtained, from which the causes affecting the 

 yield can be deduced, by parallel plotting of all the known 

 environmental changes. The continuous fruiting of cotton 

 makes it a peculiarly difficult crop to analyse without some 

 such method. The mere figures for yield mean very little, 

 since the same final result may be reached in an infinite number 

 of different ways. If we can not only ascertain exactly how the 

 yield was produced, day by day, but also trace back the fruits 

 to their origin as flowers, and the flowers to their origin as 

 buds on the scaffolding of flowering branches, we have resolved 

 the agricultural problem into components which the botanist 

 can deal with. 



It will be seen from the previous remarks that the writer's 

 bent is towards intensive work, as promising to yield the best 

 results in the future application of science to the cotton trade. 

 This raises a very big discussion : whether the support given to 

 scientific w r ork on cotton shall be devoted to " survey " work, 

 in which much is left to the intuition and subjectivity of the 

 observer, and the integrated agricultural phenomena are 

 observed as such ; or whether intensive work, arriving at 

 deductions by more minute methods, with detailed analysis of 

 phenomena, shall be the accepted system. 



There is certainly this much to be said for the intensive 

 workers, that the survey work has had a long innings. More- 

 over, while intensive work can partially replace survey, by 



