NATURAL DISSEMINATION. 89 



The fall movement of the weevil has been studied carefully each 

 year since 1904. (See fig. 21.) The circumstances have been differ- 

 ent each season, but with uniformity within certain limits. Several 

 examples will be given. In the fall dispersal of 1904 the weevils 

 seemed to have crossed the line of continuous infestation in southern 

 Louisiana about August I, and a little later toward the north, but 

 in all cases the movement had crossed the line by the 20th of August. 

 In this year there were two very well-defined dispersals with about 

 a month intervening. This might indicate that the first dispersal 

 was caused by the lack of food and that in another month a new 

 generation found itself confronted by the same conditions as its 

 predecessor and was also forced to disperse. 



In 1906 the movement seems to have been more irregular, for the 

 first serious new infestation was in central Louisiana rather than 

 in the southern part of the State. In the light of present knowledge 

 this was probably due to the smaller amount of cotton grown in the 

 pine woods of southern Louisiana, which naturally gave rise to 

 comparatively few weevils for the flight. The year 1906 was the 

 last in which any appreciable movement into western Texas was 

 observed until 1910. 



In 1907 and 1908 the eastward and northeastward progress of 

 the weevil carried it far into regions where much cotton is produced. 

 The year 1909 exhibited some very striking features. There had 

 been a considerable loss in the infestation during the winter of 1908 

 in northern Louisiana and eastern Arkansas, a region of very exten- 

 sive cultivation of cotton. During the autumn of 1909 the almost 

 continuous movement in southern Mississippi from field to field in 

 the rather sparsely cultivated areas amounted to 120 miles for the 

 season. In the delta region of Louisiana, Mississippi, and Arkansas, 

 where the weevils encountered a belt of extensive cotton culture from 

 which they had been driven back during the previous winter and 

 were stopped by the large amount of food available, they were 

 unable to gain more than 20 miles of new territory. 



In 1910 a peculiar situation developed. It was discovered that 

 high winds had caused an extensive movement into central Mississippi 

 in May or June. In the entire history of the weevil there had pre- 

 viously been known but one occasion when a severe storm caused 

 a dispersal of the insect. A study of the records of the Weather 

 Bureau brings out the fact that there was a series of cyclonic storms 

 about May 7, 1910, passing northeastward across Mississippi from 

 the heavily infested regions around Natchez. We have been unable 

 to find any other explanation of such an extensive movement in the 

 early spring. Studies conducted during the summer and fall of 1910 

 revealed the existence of many sporadic infestations throughout 

 central Mississippi, probably due to the storm. From these isolated 

 infestations the weevils spread in concentric circles until about the 

 end of November, when the intervening territory became covered. 



The winter of 1909-10 was unfavorable to the weevil in the Delta. 

 When the dispersion season opened it was noticed that in strong 

 contrast to the rapid movements in central Mississippi, the weevils 

 in the Delta advanced slowly. During the entire season there were 

 only two courses of considerable movement in the Delta region. One 

 of these was along the Mississippi River through the fields adjoining 

 the levees. The other extensive movement in the Delta country 



