48 THE DEVELOPMENT OF RAW COTTON 



ditions being constantly about 40 per cent, year in and 

 year out. This proportion varies daily (Fig. 15) with 

 the weather, and especially with the water- 

 supply, any severe water shortage increasing 

 the amount of shedding; while after each watering if 

 not excessive the proportion of shed flowers diminishes, 

 to show up again some fifty days later in the form of 

 an increased rate of bolls opening. 



It seems to have escaped notice that the shedding takes 

 place almost entirely in the flower stage. The fact that a 

 certain proportion of the sheddings (as collected from the 

 ground below the plants) consists of bolls and buds is rather 

 misleading. At any given moment there are enormously 

 more buds and bolls on the plants of a field than there are 

 flowers ; in the middle of the flowering season the ratio would 

 be about 20 buds : 1 flower : 40 bolls, simply because of the 

 differences in the duration of each stage of development. If, 

 therefore, we even found equal ratios of buds, bolls, and flowers, 

 in the sheddings, this implies that 40 per cent, of the available 

 flowers have been shed for every 1 per cent, of available bolls, 

 and for 2 per cent, of the buds available. 



The flower stage is thus extremely liable to shedding, pos- 

 sibly for reasons connected with the chemical side of pollina- 

 tion or with the greater transpiration of water from the open 

 flower. 



Apart from shedding, the only other factor which can 

 seriously influence the yield is the size of the individual 



boll, and this, again, appears to be specific 

 The Boll. , 



tor each kind of cotton under definite cir- 

 cumstances, so that cotton-breeding for a big boll would 

 appear to be a profitable line of research. With the 

 opening of the flower the purpose of the present chapter 



