32 



exceedingly small number of weevils on Brazos bottom plantations 

 in 1903 must have been due to the flood late in the Avinter preceding, 

 which undoubtedly killed large numbers of the weevils hibernating 

 on the ground below the water level in the fields or elsewhere. Had 

 this flood occurred after the emergence of the weevils it would prob- 

 ably not Jiave materially reduced their numbers. A larger number of 

 weevils always appear about outbuildings and barns and, very notice- 

 ably, near gins. 



The mortality of the hibernating wecA^ils is a matter of consider- 

 able importance, for, were it not for the large number which fail to 

 survive the winter, it woidd be impossible to grow cotton in the 

 infested region. Of 500 weevils going into hibernation in cages 

 where they were largely protected 7 per cent survived. Hunter and 

 Hinds have indicated that about 15 per cent survive at Victoria, and 

 the data available would indicate that this is usually about the per- 

 centage for southern Texas. At College Station, however, extensive 

 observations show that normally only about 2 per cent, and rarely 

 over 5 per cent, survive in the field; or, in other Avords, but one-fifth 

 as many as in southern Texas. That this is a most important factor 

 in determining the possible amount and time of damage the next 

 season is apparent. If a similar dift'erence is found betAveen central 

 and northern Texas it Avill l)e a matter of great importance for the 

 latter section of the State and other portions of the cotton belt. 

 Careful estimates of the number surviAnng in southern Texas during 

 the last Avinter shoAV that in LaA^aca County fully 30 per cent, or 

 tAvice as many as usual, survived. With this number aj^pearing in 

 the spring, amounting to about 2,500 per acre by actual count, it is 

 impossible to raise a profitable cotton croj) by any means noAV knoAvn. 

 .Vs a result, for the first time since they haA^e been infested, the 

 counties of southern Texas, Avhich haAc heretofore shoAved no marked 

 decrease in production OAving to tlie Aveevil, i)roduced almost no crop. 



The method used for determining the number of Aveevils surviving 

 \^'as to count a large number of stalks in the fall at time of hiberna- 

 tion and determine the number of Aveevils per stalk; then, in the 

 spring, to count them in a similar Avay until the first sununer brood 

 commenced to emerge. In this AA'ay tlie number of Aveevils per acre, 

 both in fall and spring, can be A^ery accurately determined, provi<led 

 large numbers of stalks are counted in several fields in one vicinity; 

 and Ave believe this to be the most accurate method of determining the 

 actual mortality Avhicli takes place in the field under natural con- 

 ditions. 



The time of the greatest mortality is a matter of some interest. As 

 nearly as can be judged from the meager data now available, the 

 largest number died in December coincident Avith the greatest rainfall 

 of the Avinter, Avhich Avas above normal for that month. Usually the 



