3l8 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I25 



larvae are found in the head region of infected wild-caught flies in 

 comparison to what occurs in the mosquito, Culicoides spp., and other 

 insect vectors of filarid worms. Actually, since the natural-infection 

 rate in Simulium spp. is itself so low, it should be expected that the 

 great majority of infected flies would contain developmental stages of 

 the filarid larvae in the thoracic muscles rather than infective forms. 

 However, a study of the feeding time of the three principal anthropo- 

 philic species does offer a plausible explanation for the paucity of 

 infective larvae in the head region of wild-caught flies. 



Numerous observations were made of flies feeding on human sub- 

 jects. The following information was recorded: Species, activity of 

 the fly on the subject prior to feeding, region of body chosen, and 

 time required for feeding. It was at first thought preferable to include 

 in the data only flies that had fed to repletion, but it was impossible 

 to choose the satiation point since many engorged flies fell off the sub- 

 ject and died. It was therefore decided to use the general feeding time 

 of all flies. In table 14 the time required for feeding indiscriminately 

 on all parts of the body has been averaged for each species. In table 

 15 is presented the feeding time of flies biting only on preferred re- 

 gions of the body (i.e., upper body region for 5". ochraceum, and lower 

 limbs for 5. metallicum and calUdum). (See also Appendix III, text 

 fig. 14.) 



It can be seen in table 14 that vS". ochraceum, metallicum, and calli- 

 dum all feed over a comparatively long period of time. The ranges of 

 feeding time for these species are: i to 19 minutes, i to 31 minutes, 

 and I to 15 minutes respectively ; the mean feeding times are 4.8, 4.3, 

 and 4.5 minutes respectively. When the feeding times on preferred 

 regions alone are considered (table 15), the means for the three spe- 

 cies are more dispersed and the differences between them are statis- 

 tically valid. In all three species, the feeding time was reduced when 

 the areas attacked contained numerous superficial blood vessels. Since 

 the bite of the black fly is usually not perceived until the feeding has 

 progressed rather far, and since the actual feeding extends over a rela- 

 tively long period of time, the infective stages in the flies would have 

 ample time to migrate from the thoracic muscles to the mouthparts, 

 from which they can then enter the wound. 



Simulium ochraceum differs from metallicum and callidum insofar 

 as its behavior on the host just prior to, and after, piercing the skin 

 is concerned. In general, ochraceum is less "nervous" than the other 

 two species. It will land on a body region without much initial hover- 

 ing, and after landing it moves about very little before piercing the 

 skin and taking blood. The mouthparts are introduced to a great 



