114 THE DEVELOPMENT OF KAW COTTON 



fluctuation it shows, even under conditions which were 



almost ideal for a field crop. Obvious senescence ap- 



pears to have been excluded entirely, but 



Fluctuating in spite of thig the ^gfo swings steadily 



Ju6ngt>n. 



over a range of 1-4 millimetres on the 



five day means, which in actual daily data of the same 

 precision would be nearly 2-5 millimetres, or T V inch. 

 The length of the lint as shown in these experiments is 

 determined by combing on the seed; if expressed as 

 the length of a "pulled tuft," in the usual way, this 

 fluctuation would be stated as a change from about If to 

 If inches, which, though not great, is still appreciable in 

 the manufacture of combed yarns. 



Lastly, it will be seen that after the lint length curve 



has been shifted 23 days to fit the strength curve, each 



irrigation shows an effect upon it. On the 



Effect of third or fourth day a f ter wa tering the two 



Irrigation. 



superposed curves begin to rise, continue to 



rise until about the tenth day after watering, and then 

 die away again, to be revived by the next irrigation. 



Adding on the 23 days of the shift, this means that 

 the length shows the first signs of having been affected 

 by watering in those bolls which open 27 or 28 days 

 afterwards, and that the maximum effect of watering 

 upon the length is shown in those bolls which open about 

 32 days afterwards. 



To obtain the age of the boll at which the effects of 

 watering are produced, we must subtract these intervals 

 from the maturation period; this period was determined 

 for this series by an indirect method as being 48 days. 



