April, '08] JOURNAL OP ECONOMIC ENTOMOLOGY 131 



it is necessary that there be from one to three generations of parasites 

 on the spring weevils, thus producing a good supply of individuals at 

 the time the boU weevil begins work. It is, however, best that the boll 

 weevil receive the earliest generation possible in order to prevent a 

 division with the other hosts. To insure the proximity of parasites 

 to the cotton field in early spring, it appears advisable to have plants 

 such as dewberry or blackberry in hedgerows, or to have red haw 

 trees near. In the first case the strawberry weevil, Anthonomus sig- 

 natus, is quite generally distributed and serves as an available host 

 for Catolaccus incertus. The Crataegus trees are the food plants of 

 three or four species of weevils which are co-hosts with the boll weevil. 



During the summer season, there is an extensive series of host 

 weevils in neighboring weeds which can be made to give up many 

 parasites, if the weeds are cut when at their height, about twice during 

 the summer season. The practice of making hay in the vicinity of 

 cotton will bring about similar results. The principle is that the 

 parasites will be forced to seek new hosts and will take the predomi- 

 nant related host — namely, the boll weevil. This is not a theory, for 

 it is substantiated by definite experiments on the cotton farm at 

 Dallas. 



Adjacent to one edge of the Dallas experimental farm was a high 

 hedge of Ambrosia trifida infested by Lixus scrohicollis, which is usu- 

 ally highly parasitized by Eurytoma tylodermatis. In 1906 this por- 

 tion of the field showed less than" three per cent parasitism due to 

 this species in hanging squares. At the time of cutting the weeds, 

 check examinations were made and two weeks later another was taken, 

 showing a considerable gain in attack by Eurytoma, which netted over 

 30 per cent. The careful records kept on this field preclude the possi- 

 bility of ascribing this result to any other cause. 



In southern Texas where the predominant trees are leguminous, any 

 cause which would tend to check the fruiting of the huisache and 

 mesquite in alternate years or at irregular periods, would tend to 

 cause an overthrow of the normal habits of the many parasites of the 

 bruchids in the pods and drive them to the boll weevil. Our atten- 

 tion was forcibly called to a particular field at Victoria with high 

 parasitism, where the presence of the new Ceramhycohius on the boll 

 weevil was first noticed and was definitely traced to the huisache trees 

 which had failed to fruit this year. 



The most important of all the cviltural suggestions for control of the 

 boll weevil is the early destruction of the cotton stalks. Owing to the 

 probability that the parasites can hibernate better when attacking the 

 native weevils, this practice seems advisable in order to drive the 



