74 NATURAL CONTROL OF THE COTTON BOLL WEEVIL. 



amounted to full}^ 70 per cent. From 10 to 15 per cent of mortality 

 occurred from other causes than those given in tlie tables. 



All factors of natural control seem to operate more effectively 

 against weevil stages in squares than against those in bolls. Ants and - 

 heat or drying are the most important factors in each class. In bolls 

 about two-thirds of the stages found were alive, while in squares 

 but two-fifths were living. 



The data used in this bulletin include examinations in twenty- 

 eight localities, in several fields in each locality, and in seventeen 

 places examinations upon from two to nine dates between June 15 

 and October 15, 1906. More than 86,000 forms were examined and 

 39,000 weevil stages were found. 



Nearly all of the eleven localities having an average total mortality 

 above the average for the twentj^-eight localities examined are situ- 

 ated south of the center of cotton production in Texas. 



Ants of the species Solenopsis geminata are more important in the 

 summer control of the weevil than are heat and parasites combined. 



The effectiveness of heat from sunshine is largeh" influenced by 

 spacing of plants, wliich should be wide for best results, and by the 

 coincident dr3niess of soil or atmosphere. The mortality from heat 

 in two groups of localities having almost identical mean maximum 

 temperatures varies as widel}^ as between 7 and 27 per cent. Exact 

 reasons for this great difference are not apparent. Average climatic 

 variations do not appear to produce a corresponding variation in the 

 average mortality of weevil stages. 



Ants seem to enter only forms containing living stages of the 

 weevil. Variations in mortality from ant attack may be due to vary- 

 ing abundance of the ants in different fields or localities. Their 

 activity is evidently influenced by climatic conditions and may also 

 be affected somewhat by cultural conditions. 



All parasites attacking the weevil have other hosts. Parasite 

 attack is much greater in hanging than it is in fallen forms. The 

 abundance of parasites seems to depend largely upon the proximit}^ 

 of some other plant than cotton in which their usual hosts abound. 

 Their attack upon tlie boll weevil is a promising adaptation to a new 

 host and may increase in effectiveness. 



CUmatic and cultural conditions seem to have been less influential in 

 producing the three highest percentages of mortality than Avere the 

 ants alone. 



The liighest mortality from heat or dr}H[ng was foimd at Corpus 

 Christi, Tex., in coincidence with comparatively low average max- 

 imimi and average mean temperatures, but after an exceptional 

 drought extending over some eight or ten weeks. This occurred 

 also in a well-tilled field where not more than one-half of the groimd 

 was shaded. 



