ii8 THE COTTON PLANT IN EGYPT chap. 



one or two suspicious cases have been recorded. The 

 method has its own disadvantages, however, the first being 

 that of expense, while the second is that some strains resent 

 the treatment, and refuse to hold their bolls. Rattoon 

 plants, i.e., plants which have been cut back and allowed to 

 shoot again in subsequent years, usually grow well under 

 the nets, as also do most American Uplands, but — though 

 erratically — ^many Egyptians are failures. When the 

 method was first employed we had yet to discover the 

 sunshine effect ; this effect is precluded from its full 

 operation by the permanent veil of netting, and netted 

 plants consequently grow during the day, becoming 

 abnormally tall. Improvements are being made by 

 substituting wire gauze for mosquito net, using larger 

 cages, and so forth, since the perfection of some preven- 

 tive method of this nature lies at the very foundation of 

 all cotton-breeding and of seed- supply. ^^ 



Another obvious possibility is the discovery, or manu- 

 facture, of a cleistogamic flower, which shall obstinately 

 refuse to admit foreign pollen to its style. At one stage 

 of these researches the author seemed to be well on the 

 road to success in this direction, and the story of the 

 ultimate failure is not without suggestiveness. 



The short-style flower. — At the time when it was 

 being realised that natural crossing would be a permanent 

 source of trouble, confusion, and error, the question of 

 floral structure naturally came under consideration. No 

 hint of the existence of uncrossable cotton flowers * could 

 be found, but it seemed reasonable to expect that if we 

 could decrease the opportunity for foreign pollen to reach 

 the style, we might expect vicinism to diminish.^' 

 v^ The cotton flower has a dense brush of anthers, borne on 

 a cylindrical column, through the centre of which the 

 style projects. The length of this style, and the extent 



* Compare Howard and Howard recently on the related genus Hibiscus. 



