52 



NATURAL CONTROL OF THE COTTON BOLL WEEVIL. 



than that found in Texas. Nearh^ all of the excess in total mortality 

 found in Texas is shown to be due to heat or drying. A sufficient 

 explanation for this difference in the effect of heat in the two States 

 is not brought out b}' the notes. B}^ reference to Table YII it may be 

 seen that only one locality in Louisiana (Mansfield) is to be found in 

 Group A, wliich contains those localities having a total mortality 

 greater than 52 per cent. In that case it is quite evident that ants 

 were responsible for about three-fourths of the total mortality found. 

 It is also evident that the mortality from heat in that locality was 

 greater than at any other localit}^ examined in Louisiana. This was 

 doubtless due to the early defoliation of the plants by rust, but even 

 in this case it was hardly equal to one-half of the average mortality 

 from heat found in Group A. In Table YIII it is sho^^Ti that all of 

 the fields examined in Louisiana had a sandy soil. This soil condi- 

 tion may have been related in some way to the larger mortality from 

 ant work and to the smaller mortality from heat or drpng, but further 

 observations are needed to definitely determine the correctness of this 

 supposition. It is noticeable that parasites are somewhat more active 

 in Texas tlian in Louisiana, and the difference shown probably indi- 

 cates fairly correctly the degree of adaptation to weevil conditions 

 which the Texas parasites have undergone. This seems especialh^ 

 true in view of the fact that a general average of observations upon 

 parasite attack made in Texas four 3"ears ago also showed 3 per cent, 

 as do the Louisiana localities in 1906. 



In comparing the figures given for each class of forms in the two 

 States some peculiar conditions are noticeable. Among hanging 

 forms in Louisiana the total mortality averaged 48.6 per cent, while 

 for the same class of forms in Texas the total mortality averaged but 

 37.9 per cent. On the other hand, for fallen forms in Louisiana the 

 average was 39.6 per cent, while in Texas it was 57.9 per cent. The 

 reason why the percentage of mortalit}' is higher in hanging forms 

 than it is in fallen forms in Louisiana while it is higher in fallen than 

 in hanging forms in Texas is not apparent. A tabidar statement of 

 these diflerences presents them most clearly : 



State and class of forms. 





Mortality caused l>v- 



Heat. 



15 5 



(>. 9 



12.0 



25.3 



Para- 

 sites. 



20.2 

 32! 2 

 16.5 

 29.7 



6.9 

 0.5 

 9.4 

 3.1 



Total 

 mortal- 

 ity. 



48.6 

 39.6 

 37.9 

 57.9 



Tlu^ conclusion that is strikingly apparent is that, for some reason, 

 the mortality among the forms on the ground is much greater in 

 Texas than it is in Louisiana from the factor of heat or dr^^ing. As 

 is shown in Table \ll, the temperature dilferences are not sufficient 



