38 PAEASITES OF THE COTTON BOLL WEEVIL. 



individual for each host. The developmental period seems to be 

 considerably shorter than in either of the species preceding. ' The 

 data upon which the following figures are based cover the period 

 from June 25 to October 28. During June and July the total develop- 

 mental period is known to be over 11 or 12 days, during August it is 

 over 14 days, and during September and October over IS day-. 

 The pupal period in June and July is only 4 to 6 days, in August 6 

 to 7 days, in September 5 to 9 days,. but in October 13 to 15 day-. 



The larva? of Catolaccus were described in the notes as having 

 short hairs in a row on each segment. The pupa? are white to yellow, 

 the thorax turning black first as they approach maturity. The 

 abdomen is very flat below and bent at an obtuse angle to the thorax. 

 The head is very broad, and the eyes are pink. The exuvium is 

 vellowish and resembles a little pointed cap. (See figure of pupa, 

 PL I, fig. 2.) 



6. Sigalphus curculionis Fitch. Xo corroboration has been recorded 

 of the parasitism of the boll weevil by this species. This parasite 

 is very commonly bred from the plum curculio (Conotrachelus nenu- 

 phar Hbst.). Chittenden also records it as a primary parasite of 

 Trichobaris trinotata Say, which breeds in the stems of Solarium 

 rostratum (Chittenden, 1902). 



7. Urosigalphus anthonomi Cwfd. This parasite was again bred 

 September 20 from material collected September 5 at Brownsville. 

 The period in the cocoon was at least 9 days. The cocoon with 

 weevil larva's head attached was half in a cotton seed in a boll. It 

 was very much finer meshed than that of Bracon mellitor and easily 

 broken. In addition to this species and (8) Urosigalphus schicarzi 

 Cwfd., both bred from Anihonomus grandis, Urosigalphus armatus 

 Ashm. has been bred from Balaninus, and another species ( Urosigalph us 

 bruchi Cwfd.) from Bruchus prosopis Lee. taken at Harlingen, Tex., 

 in beans of Prosopis glandulosa. Hence it is very probable that the 

 species of this genus are normally weevil parasites. 



9, 10. Bracon mellitor Say. Bracon dorsator Say. Bracon xantho- 

 stigma Cress. 



These are all forms of a single very variable species, of which the 

 entirely red form holds the name B. mellitor, that with black sternum 

 and metathorax the name B. xanthostigma, and that with the thorax 

 almost entirely black and the vertex of the head also black, being 

 also smaller in size, has been known as B. dorsator. 



A peculiar record is furnished by F. H. Chittenden who reared this 

 species from the strawberry leaf-roller Ancylis {Phoxoptiris) comp- 

 tana FroL, from material collected at Cabin John, Md., July 9. 1899. 



In a personal letter dated August 30, 1906, Dr. Wm. H. Ash- 

 mead writes: " Bracon mellitor Say is undoubtedly a most important 

 parasite, widely distributed in the United States, and must affect 



