116 THE DEVELOPMENT OF RAW COTTON 



If we next consider the extent of the change which the 

 strength undergoes between the effects of one watering 

 Greater an( * ^at of the next, we find that it is very 

 Fluctuation much greater than in the case of the length, 

 in Strength. T ne change which we found between 32- 7 and 

 34- 4 millimetres in the five-day means of lint Length stands 

 in the ratio of 95 : 100. In the case of lint Strength the 

 extreme values recorded stand as 65 : 100. This greater 

 capability for fluctuating is not due to accidents of 

 method, for the fact that our impact test determinations 

 are less precise than those of length should help to ob- 

 literate such distinctions. We may safely assert that in 

 this series the strength of the lint rose from 60 to 100 in 

 a single week (August 22 to 29). 



This phenomenon introduces a proposition of very 



practical interest to the spinners of fine cotton. Would 



it not be worth while to encourage the 



Short-Period practice of picking at shorter intervals ? 

 Pickings. 



If we regard these curves as an analysis of 



the three " pickings " in which the cotton crop is usually 

 harvested, it becomes clear that the composition of the 

 lint at any one picking cannot to uniform, except by 

 accident against long odds. In conventional practice 

 the bolls which we have studied here would have been 

 gathered in three groups or pickings, on or about Sep- 

 tember 10, September 30, and October 20. By reference 

 to the curves it will be seen that pickings taken on any 

 of these three dates would have been irregular in length 

 and in strength, since they would consist of all the bolls 

 ripened before that date. In general experience the first 



