CACTUS-FEEDmG INSECTS AND MITES 35 



The incubation period in September and October occupies 12 to 

 19 days. The young larvae develop slowly during winter; active feeding 

 commences in April. Much difficulty was encountered in attempting to 

 rear this insect in cages at Uvalde, Texas. The young larvae survived 

 the winter months, but invariably succumbed to disease in the spring 

 and summer. Large larve, collected from the field and housed at Uvalde 

 in July and August, often suffered the same fate. Similarly, efforts to 

 establish the species in Australia were abortive, since the larvae died 

 rapidly during humid weather. M. doddalis inhabits regions having 10 

 to 15 inches of rainfall; the annual precipitation at Uvalde and at the 

 Australian rearing stations is approximately 25 inches. 

 Parasites. — Pupal parasites include the tachinids Phorocera texana and 

 P. comstocki; the ichneumons Temelucha sinuatus, T.facilts, and Chelonus 

 electus; and the chalcid Brachymeria {Pseudobrachymeria) pedalis. The 

 braconid Bracon hebetor attacks the larvae of this and other cactus 

 phycitids in cages. 



Melitara Walker, sp. 



This form from southern Texas has been known to the Board's Officers 

 as Melitara bollii (Zeller). Heinrich states that the true bollii is a synonym 

 of M. prodenialis Walker, and in his 1939 monograph included bollii 

 nndcT prodenialis. In a letter (1939), however, he expressed the opinion 

 quoted below, which he further amplified in his 1956 monograph (p. 

 241): 



Recently Mr, Leith Hitchcock sent us a large number of specimens 

 representing the rearing experiments in connection with your prickly-pear 

 investigations. Among them was a series from Uvalde, Texas, labelled 

 Melitara bollii and evidently what you wanted to separate from prodenialis 

 as at least a distinct race. I understand now your position. They are part of 

 what Dyar included under his doddalis and which I have identified as 

 dentata Grote (equals doddalis) from various Texas localities. They are in fact 

 only a south Texas form of Grote's species. I do not think that we are justified 

 (with our present knowledge) in separating the south Texas specimens 

 from dentata even as a local race; but in any event the name bollii could 

 not be applied to them. 



The recognition of this insect as a distinct species or race is based 

 on the fact that it differs from M. prodenialis, M. dentata, and M. doddalis 

 in one interesting larval habit and to a lesser degree in its choice of 

 pupation sites. On hatching from the eggs in October and November, 

 the larvae from one eggstick form a small cell, of one-half inch to one 

 inch diameter, near the margin (usually the upper margin) of the 

 prickly pear segment. This cell is indicated externally by the dead 

 cuticle. The larvae remain in the cell throughout winter and must 



