BREEDING IN BUDS A DERIVED HABIT. 63 



also that if the weevil fed originally upon the bolls it has followed 

 hack to earlier and earlier stages and finally to the bud. The facts 

 already detailed seem to prove, however, that this is not the case. 

 The weevil does not attack the very young bolls, nor does it operate 

 while the flower is open or while it remains in place, though in a 

 withered condition. The hatching of the weevil larva in the large 

 buds is likewise inetfective because the larva is deprived of shelter 

 when the flower opens. It seems necessary to believe, therefore, that 

 the parasitism of the weevils upon the buds of the cotton is a habit 

 (juite distinct from that of its relations to the boll. The habit of 

 breeding in the bud marked a new departure in the biological history 

 of the insect and not a gradual change from the previous habit of 

 infesting the bolls only. Nevertheless, the change of habits need not 

 be thought of as anything very remarkable from the standpoint of the 

 insect. A cotton bud is very nuich larger than a small boll. The 

 peculiarity lies in the plant rather than in the insect, since very few 

 plants afford a continuous and abundant succession of large, pollen- 

 tilled buds. It is this quality of the cotton plant which has enabled 

 the weevil to develop its peculiar and highly destructive secondary 

 haljits of feeding upon the buds and using them as breeding places. 

 If the boll weevil Avere restricted, like related beetles, to parasitism 

 upon the fruit of the cotton, it would have remained a comparatively 

 iiarmless and agriculturally insignificant enemy. These considera- 

 tions may assist in a better appreciation of the extent to which the 

 weevil's power of injury Avould be diminished if we could obtain a 

 variety of cotton with a fully determinate habit of growth, one which 

 would cease producing buds as soon as a crop of cotton had been set. 

 The much more rapid development of w^eevil larvae in the Inid is 

 to be connected, doubtless, Avith the much richer food offered by the 

 mass of poll-en, but it may represent also a somewhat more definitely 

 adaptive specialization of the life history of the weevil, for it is gen- 

 erally a question of eating the pollen })romptly or not at all. If the 

 bud falls off' on moist ground the pollen would be completely decom- 

 posed long before the larva could develop, at the rate at which it 

 grows in the boll, and if the bud did not drop off', but continued to 

 grow, the flower would open and turn the larva out. It is obliged, 

 therefore, to do damage fast enough to keep the flower from opening, 

 and must then eat the remaining ])ollen before it spoils and leaves 

 the hu-va too hungry and stunted to pass through the final meta- 

 niorphosis into the adult stage. In a cotton which has a highly 

 developed habit of shedding the injured buds it would not be so neces- 

 sary for the larva to attack the pistil. It may be that this policy 

 ou the part of the weevils in Guatemala has a use to the weevil as 

 being necessary to prevent the opening of the flower and cause the 

 falling of the bud. 



