June, '13] TOWXSEXD: COTTON SQUARE WEEVIL 309 



that they can be absent from the Casma valley. It appears more 

 likely that the abnormally humid hot season of 1912 in the central 

 coast region was mifavorable to the parasites of the weevil, but not 

 so to the weevil itself. 



On July 9, 1912, 143 yellowed and withered squares were collected 

 at Santa Beatriz, near Lima, from which were secured thirty-nine 

 weevils, nine known parasites of the weevil and five doubtful parasites. 

 The fourteen parasites secured from this lot are as follows: — 



CatoJaccus sp. probably C. townsendi^Two females and five males 

 agreeing with this species in color. 



Microbracon sp. near M. vestiticida — Two males. Differs from this 

 species in the disposition of black, but maj* be only a variety. 



Braconid sp.— One. Doubtfully parasitic on the weevil. 



Braconid sp. — Three males; and one female that seems to belong 

 with them, but differs in coloration. Rather small for parasites of the 

 weevil. 



In Texas a small black stinging ant, Solenopsis geminata, is well 

 known as a very effective enemy of the boll-weevil, often killing 25 

 or 30 per cent of the early stages in the fallen squares. This species 

 exists in the western foothills of the Andes, and it is almost certain 

 that it would destroy the early stages of the picudo in Peru as 

 effectively as it does those of the boll-weevil in Texas. A closely 

 allied ant, doubtfullj'- determined as Solenopsis pylades by Dr. Wheeler, 

 occurs commonly in the Piura cotton-fields where it is known as 

 hormiga picador, but it has not as yet been observed to prey upon the 

 weevil. Colonies of 8. geminata should be transferred from the Andean 

 foothills to the cotton districts of the coast region, and established 

 there. 



In Guatemala an ant known as the kelep, Edatomma tuhercidatum, 

 is known as a very effective enemy of the adult boll-weevils, which it 

 seeks and attacks in the green squares while they are engaged in feed- 

 ing and ovipositing. This species exists in the eastern foothills of the 

 Andes, being known to range from Guatemala south through Costa 

 Rica and the Andean montam^a to the valley of the Rio Beni in Bolivia. 

 Colonies of it should be transferred from the montanya to the cotton 

 districts of the coast region, and there established if possible. Its 

 introduction was attempted in Texas, but the conditions there were 

 unfavorable to it. The conditions in the Peruvian coast region would 

 probably prove much more favorable to the kelep than those of Texa's. 



CuLTUEAL Control Measures 



The carrxdng out of a carefully planned series of experiments is ur- 

 gently needed to demonstrate the details of cultural control measures 



