48 PARASITES OF THE COTTON BOLL WEEVIL. 



2 5. Bruchus exiguus Horn breeds in the seed pods of Amorpha 

 fruticosa in the fall. It is a host of Cerambycobius cyaniceps and 

 Catolaccus incertus. 



It is noticeable from the foregoing that Bracon meUitor attacks 

 bud, flower, pod, and seed weevils; Catolaccus incertus attacks bud, 

 pod, and seed weevils; Cerambycobius cyaniceps attacks bud, pod, 

 and seed, but mainly stem weevils; Eurytoma tylodermatis attacks bud, 

 flower, pod, and stem weevils ; Microdontomerus anthonomi attacks bud, 

 capsule, and stem weevils; Myiophasia senea attacks bud, fruit, pod, 

 nut, and stem weevils, and Sigalphus curculionis attacks bud, fruit, 

 pod, and stem weevils. Such general habits indicate a generaliza- 

 tion of habit in the species concerned and give promise of the 

 possibility of more extensive adaptation to the boll weevil as a host. 



ROTATION OF HOSTS. 



The most important item in the solution of the parasite question 

 is probably concerned with the treatment of the other hosts. It is 

 found that the parasites which now attack the weevil are naturally 

 parasitic on weevils of short season and few generations, and the 

 most feasible proposition is that of forcing them to attack the boll 

 weevil only. 



Two lines of action are suggested as possible and practicable. 

 The first is, that such plants as the Crotons, hosts of Anthonomus 

 albopilosAus , might be planted in abundance in the pastures because 

 they make good feed and are easily destroyed, and that thus the 

 weevils and their parasites would increase in great numbers and the 

 latter might perhaps attack the boll weevil. 



The counter proposal is that the timely elimination of the summer 

 series of host plants would cut off the possibility of finding any host 

 but the boll weevil, which would be in abundance. The cutting of 

 the weeds around the fields could not result disadvantageously, as 

 there are numerous pests harbored by these same weeds. At present 

 the information as to the rotation of hosts is incomplete, and it is 

 therefore necessary to go very fully into this question as well as to 

 test both methods which have been suggested. 



The known hosts of Bracon mellitor would insure it an uninter- 

 rupted rotation if they all occurred in a given locality. They do not, 

 so there are still other unfound hosts of this species. The season of 

 its hosts as far as known are as follows: 



Anthonomus fulv us breeds from May 3 to July 12. 

 Desmoris scajmlis breeds from July 25 to September 22. 

 Anthonomus squamosus breeds from AugUBl LI to September 19. 

 Anthonomus albopilosus breeds from August 26 to September 29. 

 Anthonomus eugenii breeds from October 6 to October 31. 



