COTTON-GROWING 163 



Fortunately it is a tolerably simple matter to obtain 

 such records, given a few native labourers and a little 



training. The observations may be taken 

 Crop Records. 



on the entire plots, or on groups of 200 



plants each, counted and marked off by stakes in each 

 plot. The observations which are practicable are those 

 of boiling and flowering, and the object of these observa- 

 tions, as well as of others which are less easily obtained, 

 is to present a continuous record of the behaviour of the 

 plants. The ideal would be to take these records daily 

 (Figs. 4 and 10), but this would rarely be practicable in 

 early development work, and slightly longer intervals 

 may be substituted. 



The records of boiling are obtained by picking or count- 

 ing the number of bolls open on the observed groups of 

 plants each week, so as to obtain the number 

 ( g ripening in each week per plant or per plot 

 (Figs. 10, 13, and 16). When a large number 

 of plots is being handled, it may be convenient to take 

 one day of the week for one series, and one for another, 

 but adhering strictly to the same day for each series. The 

 figures may be expressed as the number of bolls ripening, 

 or the weight thereof, or best of all both ways ; if both 

 are taken, the average boll weight each week is thus 

 obtained, which is an important consideration. By 

 plotting the results on squared paper, it is easy to see, 

 not only which plots gave the best yield, but which were 

 earliest, etc., and to deduce from these curves the reasons 

 for failures. 



The flowering records (Figs. 2 and 10) are even more 



