64 THE DEVELOPMENT OF RAW COTTON 



tubes require further study. The style of some kinds of 

 cotton is either non-nutritious, or more probably poisonous, 

 to the pollen of other kinds. Thus, crosses between the 

 Indian group of cottons and the Upland or Peruvian groups 

 do not appear to be possible. Uplands and Peruvians can, 

 on the other hand, be artificially intercrossed with ease. 

 Even in this case, however, if equal amounts of self and 

 foreign pollen are placed on the style simultaneously, so that 

 both have an equal chance, 97 per cent, of the victors will 

 be self -tubes; so that, although Egyptian pollen can grow 

 down an Upland style quite satisfactorily, it cannot grow so 

 fast as the Upland's own pollen can do, and vice versa. If, 

 lastly, the pollen mixture is made with pollen from the first 

 cross between Upland and Egyptian, the percentage of wins 

 credited to the home team falls to about 60 per cent.; the 

 hybrid pollen is said to be more "prepotent." These facts 

 have considerable economic bearing on the possibility of 

 keeping cotton strains pure with fewer precautions, but it will 

 take a great deal of tedious research to find whether they have 

 any utility. 



THE FRUIT: GENERAL (Fig. 13). The few days im- 

 mediately preceding flowering, and the day of flowering 

 itself in particular, are extremely critical ones in the 

 history of the fruit. For some reasons which are not yet 

 understood, the open flower is extremely liable to 

 "shedding" (see p. 48). The cause which provokes 

 such shedding, with consequent complete loss of the fruit 

 from its corner of the plant scaffolding, is usually shortage 

 of water, either for the whole plant or for the particular 

 fruit concerned. The shedding of older fruits is much 

 rarer, and in all probability the disposition to shedding 

 is connected with the presence of some chemical sub- 

 stance or substances formed just at the flowering stage. 



