66 INSECT EI^EMIES OF THE BOLL WEEVIL. 



About the middle of August squares commence to fail, and few 

 squares are to be found by September 1. This condition may be 

 said to begin the fall season, when the parasites are largely obliged 

 to seek other hosts or to attack the boll weevil in bolla. 



IV. THE PALL OR DISPERSION SEASON. 



Coincident with the decline in square production is the beginning 

 of the boll-weevil dispersion which extends into new territory around 

 the entire periphery of the infested region. In the fall there is a new 

 growth of squares which furnishes food for the weevils before entering 

 hibernation and also furnishes an opportunity for very high parasitism 

 just preceding hibernation. It is during this season that parasite 

 swarms are recorded and hence this is a very critical time for obtaining 

 and transferring desirable parasites to new regions. During this early 

 faU season there are several very important ways of propagating the 

 parasites already present in the vicinity, as will be shown later. The 

 faU season of the year closes abruptly mth the first killing frost, for 

 this crisis precipitates the hibernation period. 



10. ADJUSTMENT TO NEW HOSTS. 



It is a very striking fact that the continuous!}^ breeding boU 

 weevil is attacked by parasites which in man}^ instances attack nor- 

 mally weevils having but a single generation annually. Some of these 

 parasites attack one host after another throughout the entire breeding 

 season and may be found in activity at all periods except during 

 hibernation. This condition is weU illustrated by the accompanying 

 diagram (fig. 15) giving the seasonal rotation of Catolaccus Jiunteri 

 and Ceramhycobius cyaniceps. Wliether these parasites were origi- 

 naUy single-generation species like their hosts is a question we can not 

 now decide, but we now know that they have become adapted to 

 many species. This fact can be most easily proven by reference to 

 the list of hosts of the boll- weevil parasites given in the second section 

 of this part (p. 42). It appears possible that the constantly changing 

 factors of nature cause the various species to be continually adjusting 

 their habits to new environments and new hosts. In other words, the 

 groups of parasites from which the most available enemies of a new or 

 introduced species may be obtained are those groups in which the 

 parasitic habits are the most variable. A parasitic species that is as 

 readily at home on a stem weevil as on a bud or seed weevil is probably 

 able to attack many different species. 



The most striking example of the adjustment of new parasites was 

 furnished in 1907. A lot of hanging squares collected by ^Ir. J. D. 

 Mitchell on August 5, 1907, at Victoria, Tex., on a field known as the 

 Haskell field gave a percentage of 61.5. There was something so 



