ENVIRONMENT OF THE BOLL 99 



alterations in general slope, which are due to the self- 

 poisoning effect. 



This identity of the two curves fits in perfectly with 

 the microscopic evidence, and we shall see in the next 

 series that it was no mere accident of the season which 

 brought it about, nor any abnormality of the plots, but 

 that it is even more marked and definite in good cultiva- 

 tion than in experimental modifications. 



Before considering the behaviour of the other characters 

 it may be worth while to comment on one detail of the 

 preceding pages. The author has shown elsewhere that 

 it is possible to depict " good cultivation " graphically, 

 for under a given set of climatic circumstances a given 

 kind of cotton should flower and fruit at certain definite 

 rates, and if these rates are not attained cultivation has 

 been defective. Following on from this, and from the 

 remarks on poisoning from water-shortage just made, 

 and supplementing them by comparison with the series 

 of dated bolls in field cultivation, it becomes possible to 

 define the object of good cotton cultivation as a fight 

 against self -poisoning, or senescence, or autotoxy. 



Seed- weight is a feature which we can hardly expect 

 to dissolve into its components. To some extent it is 

 determined by the size of the seed, which is 

 settled at the same time as lint length, and 

 should therefore fluctuate with it. The mere size, how- 

 ever, is not everything, and all the subsequent changes 

 which the embryo and seed-coats undergo must each 

 leave its mark upon the weight finally attained. We 



